Article

A 62-year-old man presented to the emergency department (ED) with 3 days of chills, myalgias, and nausea. The patient’s oral temperature at home ranged from 99.9 to 100.1 °F. He came to the ED after multiple phone discussions with primary care nursing over 3 days. His medical history included posttraumatic stress disorder, enlarged prostate, osteoporosis, gastroesophageal reflux, glaucoma, and left eye central retinal vein occlusion. Medications included fluoxetine 20 mg twice daily, omeprazole 20 mg twice daily, rosuvastatin 10 mg once daily, tamsulosin 0.4 mg nightly, and zolpidem 10 mg nightly. The patient’s glaucoma had been treated with a dexamethasone intraocular implant about 90 days earlier. The patient started on intravenous (IV) zoledronic acid for osteoporosis, with the first infusion 5 days prior to presentation.

In the ED, the patient’s temperature was 98.2 °F, blood pressure was 156/76 mm Hg, pulse was 94 bpm, respiratory rate was 16 breaths per minute, and 98% oxygen saturation on room air. He was in no acute distress, with an unremarkable physical examination reporting no abnormal respiratory sounds, no arrhythmia, normal gait, and no focal neurologic deficits. A comprehensive metabolic panel was unremarkable, creatine phosphokinase was 155 U/L (reference range, 30-240 U/L), and the complete blood count was notable only for an elevated white blood count of 15.3 × 10⁹/L (reference range, 4.0-11.0 × 10⁹/L), with 73.4% neutrophils, 16.2% lymphocytes, 9.1% monocytes, 0.5% eosinophils, and 0.4% basophils. The patient’s urinalysis was unremarkable.

What is your diagnosis? 

How would you treat this patient?

The ED physician considered viral infection and tested for both influenza and COVID-19. Laboratory results eliminated urinary tract infection and rhabdomyolysis as possible diagnoses. An acute phase reaction to zoledronic acid was determined to be the most likely cause. The patient was treated with IV saline in the ED, and acetaminophen both in the ED and at home.

Although initial nursing triage notes document consideration of acute phase reaction to zoledronic acid, the endocrinology service, which had recommended and arranged the zoledronic acid infusion, was not immediately notified of the reaction. It does not appear any treatment (eg, acetaminophen) was suggested, only that the patient was given advice this may resolve over 3 to 4 days. When he was seen 2 months later for an endocrinology follow-up appointment, he reported that all symptoms (chills, myalgias, and nausea) resolved gradually over 1 week. Since then, he has felt as well as he did before taking zoledronic acid. However, the patient was wary of further zoledronic acid, opting to defer deciding on a second dose until a future appointment.

Prior to starting zoledronic acid therapy, the patient was being treated for vitamin D deficiency. Four months prior to infusion, his 25-hydroxyvitamin D level was 12.0 ng/mL (reference range, 30 to 80 ng/mL). He then started taking cholecalciferol 100 mcg (4000 IU) daily. Eight days prior to infusion his 25-hydroxyvitamin D level was 29.5 ng/mL.

Federal health care practitioners, especially those working in the Veterans Health Administration (VHA), will commonly encounter patients similar to this case. Osteoporosisis is common in the United States with > 10 million diagnoses (including > 2 million men) and in VHA primary care populations.^1,2 Zoledronic acid is a frequently prescribed treatment, appearing in guidelines for osteoporosis management.^3-5

The acute phase reaction is a common adverse effect of both oral and IV bisphosphonates, although it’s substantially more common with IV bisphosphonates such as zoledronic acid. This reaction is characterized by flu-like symptoms of fever, myalgia, and arthralgia that occur within the first few days following bisphosphonate administration, and tends to be rated mild to moderate by patients.⁶ Clinical trial data from > 7000 women with postmenopausal osteoporosis found that 42% experienced ≥ 1 acute phase symptom following the first infusion (fever was most common, followed by musculoskeletal symptoms and gastrointestinal symptoms), compared with 12% for placebo. Incidence decreases with each subsequent infusion.⁷ Risk factors for reactions include low 25-hydroxyvitamin D levels,^8,9 no prior bisphosphonate exposure,⁹ younger age (aged 64-67 years vs 78-89 years),⁷ lower body mass index,¹⁰and higher lymphocyte levels at baseline.¹¹ While most cases are mild and self-limited, severe consequences have been noted, such as precipitation of adrenal crisis.^12,13 Additionally, more prolonged bone pain, sometimes quite severe, has been rarely reported with bisphosphonate use. However, it’s unclear whether this represents a separate adverse effect or a more severe acute phase reaction.⁶

The acute phase reaction is a transient inflammatory state marked by increases in proinflammatory cytokines such as C-reactive protein, interleukin-6, and tumor necrosis factor-α. Proposed mechanisms include: (1) inhibition of farnesyl pyrophosphate synthase, an enzyme of the mevalonate pathway, resulting inactivation of γϐ T cells and increased production of proinflammatory cytokines; (2) inhibition of the suppressor of cytokine signalling-3 in the macrophages, resulting in cessation of the suppression in cytokine signaling; or (3) negative regulation of γϐ T-cell expansion and interferon-c production by low serum 25-hydroxyvitamin D concentrations.¹¹

Prevention

Can an acute phase reaction to zoledronic acid be prevented? Bourke and colleagues reported that baseline calcium and/or vitamin D intake do not appear to affect rates of acute phase reaction in data pooled from 2 trials of zoledronic acid in postmenopausal women.¹⁴ However, patients receiving zoledronic acid had 25-hydroxyvitamin D values > 20 ng/mL 86% of the time, and values > 30 ng/mL 36% of the time. Bourke and colleagues suggest that “coadministration of calcium and vitamin D with zoledronate may not be necessary for individuals not at risk of marked vitamin D deficiency.”¹⁴ However, they did not prospectively test this hypothesis.

In our patient, vitamin D deficiency had been identified and treated, nearly achieving 30 ng/mL. The 2020 guidelines for postmenopausal osteoporosis recommend maintaining serum 25-hydroxyvitamin D levels 30 to 50 ng/mL, advising to supplement with vitamin D₃ as needed.⁵ The 2012 guidelines for osteoporosis in men from the Endocrine Society suggest that men with low vitamin D levels receive vitamin D supplements to raise the level > 30 ng/ml.⁴

Oral analgesics have been studied for the prevention of adverse effects related to zoledronic acid. Initiating 650 mg acetaminophen 45 minutes before zoledronic acid infusion and then every 6 hours over the next 3 days has been shown to significantly reduce symptoms.¹⁵ Acetaminophen or ibuprofen given every 6 hours for 3 days (starting 4 hours after zoledronic acid infusion) has been shown to reduce fever and other symptoms.¹⁶

Statins have been shown in vitro to prevent bisphosphonate-induced γϐ T cell activation.¹⁷ This has led to studies with various statins, although none have yet shown benefit in vivo. A double-blind, randomized, placebo-controlled trial of postmenopausal women for fluvastatin (single dose of 40 mg or 3 doses of 40 mg, each 24 hours apart) did not prevent acute phase reaction symptoms, nor did it prevent zoledronic acid-induced cytokine release.¹⁷ Rosuvastatin 10 mg daily starting 5 days before zoledronic acid treatment and taken for a total of 11 days did not show any difference in fever or pain.¹⁸ A protocol for pravastatin has been disseminated, but no study results have been published yet.¹⁹

Prophylactic dexamethasone has also been studied. A randomized double-blind, placebo-controlled trial of oral dexamethasone 4 mg at the time of first infusion of zoledronic acid found no significant difference in temperature change or symptom score over the following 3 days.²⁰ Chen and colleagues compared the efficacy of acetaminophen alone vs acetaminophen plus dexamethasone over several days.²¹ Acetaminophen 500 mg was given on the day of infusion and 4 times daily for 3 to 7 days for both groups, while dexamethasone 4 mg was given for 3 to 7 days. The dexamethasone group reported substantially lower incidence of any acute phase reaction symptoms (34% vs 67%, P = .003). A more recent study by Murdoch and colleagues comparing dexamethasone (4 mg daily for 3 days with the first dose 90 minutes before zoledronic acid infusion) with placebo found that the dexamethasone group had a statistically significant lower mean temperature change and acute phase reaction symptom score.²²

Adverse Effect Treatment

Treatment after development of acute phase reaction due to zoledronic acid infusion is generally limited to supportive care and/or nonsteroidal anti-inflammatory drugs (NSAIDs) acetaminophen or dexamethasone, largely based on extrapolation of the noted preventive trials and expert opinion.^3,6 Experiencing an acute phase reaction may portend better fracture risk reduction from zoledronic acid, although there is a potential association between acute phase reaction and mortality risk.^23,24

Our case was typical for acute phase reaction to zoledronic acid. The patient was already taking rosuvastatin 10 mg daily for hypercholesterolemia as prescribed by his primary care physician. Rosuvastatin was not shown to prevent symptoms, although it was not studied in patients on long-term statin therapy at the time of zoledronic acid infusion.¹⁸ The patient was also taking vitamin D₃ supplementation and was nearly in the reference range.⁵ His ED treatment included IV fluids and acetaminophen. Pretreatment (prior to or at the time of zoledronic acid infusion) with acetaminophen or ibuprofen may have prevented his symptoms, or at least lessened them to the point that an ED visit would not have resulted. The endocrinologist who prescribed the zoledronic acid documented a detailed discussion of the adverse effects of zoledronic acid with the patient, and the initial nursing call documents consideration of acute phase reaction. It is unclear whether the persistence of symptoms or worsening of symptoms ultimately led to the ED visit. Because no treatment was offered, it is unknown whether earlier posttreatment with acetaminophen, ibuprofen, or dexamethasone might have prevented his ED visit.

Conclusions

Clinicians who treat patients with osteoporosis should be aware of several key points. First, acute phase reaction symptoms are common with bisphosphonates, especially zoledronic acid infusions. Second, the symptoms are nonspecific but should have a suggestive time course. Third, dexamethasone may be partially protective, but based on the various trials discussed, it likely needs to be given for multiple days (instead of a single dose on the day of infusion). Given that acetaminophen and NSAIDs also seem to be protective (when given for multiple days starting on the day of infusion), both have lower overall adverse effect profiles than dexamethasone, consideration may be given to using either of these prophylactically.⁶ Dexamethasone could then be prescribed if symptoms are severe or persistent despite the use of acetaminophen or NSAIDs.

A 62-year-old man presented to the emergency department (ED) with 3 days of chills, myalgias, and nausea. The patient’s oral temperature at home ranged from 99.9 to 100.1 °F. He came to the ED after multiple phone discussions with primary care nursing over 3 days. His medical history included posttraumatic stress disorder, enlarged prostate, osteoporosis, gastroesophageal reflux, glaucoma, and left eye central retinal vein occlusion. Medications included fluoxetine 20 mg twice daily, omeprazole 20 mg twice daily, rosuvastatin 10 mg once daily, tamsulosin 0.4 mg nightly, and zolpidem 10 mg nightly. The patient’s glaucoma had been treated with a dexamethasone intraocular implant about 90 days earlier. The patient started on intravenous (IV) zoledronic acid for osteoporosis, with the first infusion 5 days prior to presentation.

In the ED, the patient’s temperature was 98.2 °F, blood pressure was 156/76 mm Hg, pulse was 94 bpm, respiratory rate was 16 breaths per minute, and 98% oxygen saturation on room air. He was in no acute distress, with an unremarkable physical examination reporting no abnormal respiratory sounds, no arrhythmia, normal gait, and no focal neurologic deficits. A comprehensive metabolic panel was unremarkable, creatine phosphokinase was 155 U/L (reference range, 30-240 U/L), and the complete blood count was notable only for an elevated white blood count of 15.3 × 10⁹/L (reference range, 4.0-11.0 × 10⁹/L), with 73.4% neutrophils, 16.2% lymphocytes, 9.1% monocytes, 0.5% eosinophils, and 0.4% basophils. The patient’s urinalysis was unremarkable.

What is your diagnosis? 

How would you treat this patient?

The ED physician considered viral infection and tested for both influenza and COVID-19. Laboratory results eliminated urinary tract infection and rhabdomyolysis as possible diagnoses. An acute phase reaction to zoledronic acid was determined to be the most likely cause. The patient was treated with IV saline in the ED, and acetaminophen both in the ED and at home.

Although initial nursing triage notes document consideration of acute phase reaction to zoledronic acid, the endocrinology service, which had recommended and arranged the zoledronic acid infusion, was not immediately notified of the reaction. It does not appear any treatment (eg, acetaminophen) was suggested, only that the patient was given advice this may resolve over 3 to 4 days. When he was seen 2 months later for an endocrinology follow-up appointment, he reported that all symptoms (chills, myalgias, and nausea) resolved gradually over 1 week. Since then, he has felt as well as he did before taking zoledronic acid. However, the patient was wary of further zoledronic acid, opting to defer deciding on a second dose until a future appointment.

Prior to starting zoledronic acid therapy, the patient was being treated for vitamin D deficiency. Four months prior to infusion, his 25-hydroxyvitamin D level was 12.0 ng/mL (reference range, 30 to 80 ng/mL). He then started taking cholecalciferol 100 mcg (4000 IU) daily. Eight days prior to infusion his 25-hydroxyvitamin D level was 29.5 ng/mL.

Federal health care practitioners, especially those working in the Veterans Health Administration (VHA), will commonly encounter patients similar to this case. Osteoporosisis is common in the United States with > 10 million diagnoses (including > 2 million men) and in VHA primary care populations.^1,2 Zoledronic acid is a frequently prescribed treatment, appearing in guidelines for osteoporosis management.^3-5

The acute phase reaction is a common adverse effect of both oral and IV bisphosphonates, although it’s substantially more common with IV bisphosphonates such as zoledronic acid. This reaction is characterized by flu-like symptoms of fever, myalgia, and arthralgia that occur within the first few days following bisphosphonate administration, and tends to be rated mild to moderate by patients.⁶ Clinical trial data from > 7000 women with postmenopausal osteoporosis found that 42% experienced ≥ 1 acute phase symptom following the first infusion (fever was most common, followed by musculoskeletal symptoms and gastrointestinal symptoms), compared with 12% for placebo. Incidence decreases with each subsequent infusion.⁷ Risk factors for reactions include low 25-hydroxyvitamin D levels,^8,9 no prior bisphosphonate exposure,⁹ younger age (aged 64-67 years vs 78-89 years),⁷ lower body mass index,¹⁰and higher lymphocyte levels at baseline.¹¹ While most cases are mild and self-limited, severe consequences have been noted, such as precipitation of adrenal crisis.^12,13 Additionally, more prolonged bone pain, sometimes quite severe, has been rarely reported with bisphosphonate use. However, it’s unclear whether this represents a separate adverse effect or a more severe acute phase reaction.⁶

The acute phase reaction is a transient inflammatory state marked by increases in proinflammatory cytokines such as C-reactive protein, interleukin-6, and tumor necrosis factor-α. Proposed mechanisms include: (1) inhibition of farnesyl pyrophosphate synthase, an enzyme of the mevalonate pathway, resulting inactivation of γϐ T cells and increased production of proinflammatory cytokines; (2) inhibition of the suppressor of cytokine signalling-3 in the macrophages, resulting in cessation of the suppression in cytokine signaling; or (3) negative regulation of γϐ T-cell expansion and interferon-c production by low serum 25-hydroxyvitamin D concentrations.¹¹

Prevention

Can an acute phase reaction to zoledronic acid be prevented? Bourke and colleagues reported that baseline calcium and/or vitamin D intake do not appear to affect rates of acute phase reaction in data pooled from 2 trials of zoledronic acid in postmenopausal women.¹⁴ However, patients receiving zoledronic acid had 25-hydroxyvitamin D values > 20 ng/mL 86% of the time, and values > 30 ng/mL 36% of the time. Bourke and colleagues suggest that “coadministration of calcium and vitamin D with zoledronate may not be necessary for individuals not at risk of marked vitamin D deficiency.”¹⁴ However, they did not prospectively test this hypothesis.

In our patient, vitamin D deficiency had been identified and treated, nearly achieving 30 ng/mL. The 2020 guidelines for postmenopausal osteoporosis recommend maintaining serum 25-hydroxyvitamin D levels 30 to 50 ng/mL, advising to supplement with vitamin D₃ as needed.⁵ The 2012 guidelines for osteoporosis in men from the Endocrine Society suggest that men with low vitamin D levels receive vitamin D supplements to raise the level > 30 ng/ml.⁴

Oral analgesics have been studied for the prevention of adverse effects related to zoledronic acid. Initiating 650 mg acetaminophen 45 minutes before zoledronic acid infusion and then every 6 hours over the next 3 days has been shown to significantly reduce symptoms.¹⁵ Acetaminophen or ibuprofen given every 6 hours for 3 days (starting 4 hours after zoledronic acid infusion) has been shown to reduce fever and other symptoms.¹⁶

Statins have been shown in vitro to prevent bisphosphonate-induced γϐ T cell activation.¹⁷ This has led to studies with various statins, although none have yet shown benefit in vivo. A double-blind, randomized, placebo-controlled trial of postmenopausal women for fluvastatin (single dose of 40 mg or 3 doses of 40 mg, each 24 hours apart) did not prevent acute phase reaction symptoms, nor did it prevent zoledronic acid-induced cytokine release.¹⁷ Rosuvastatin 10 mg daily starting 5 days before zoledronic acid treatment and taken for a total of 11 days did not show any difference in fever or pain.¹⁸ A protocol for pravastatin has been disseminated, but no study results have been published yet.¹⁹

Prophylactic dexamethasone has also been studied. A randomized double-blind, placebo-controlled trial of oral dexamethasone 4 mg at the time of first infusion of zoledronic acid found no significant difference in temperature change or symptom score over the following 3 days.²⁰ Chen and colleagues compared the efficacy of acetaminophen alone vs acetaminophen plus dexamethasone over several days.²¹ Acetaminophen 500 mg was given on the day of infusion and 4 times daily for 3 to 7 days for both groups, while dexamethasone 4 mg was given for 3 to 7 days. The dexamethasone group reported substantially lower incidence of any acute phase reaction symptoms (34% vs 67%, P = .003). A more recent study by Murdoch and colleagues comparing dexamethasone (4 mg daily for 3 days with the first dose 90 minutes before zoledronic acid infusion) with placebo found that the dexamethasone group had a statistically significant lower mean temperature change and acute phase reaction symptom score.²²

Adverse Effect Treatment

Treatment after development of acute phase reaction due to zoledronic acid infusion is generally limited to supportive care and/or nonsteroidal anti-inflammatory drugs (NSAIDs) acetaminophen or dexamethasone, largely based on extrapolation of the noted preventive trials and expert opinion.^3,6 Experiencing an acute phase reaction may portend better fracture risk reduction from zoledronic acid, although there is a potential association between acute phase reaction and mortality risk.^23,24

Our case was typical for acute phase reaction to zoledronic acid. The patient was already taking rosuvastatin 10 mg daily for hypercholesterolemia as prescribed by his primary care physician. Rosuvastatin was not shown to prevent symptoms, although it was not studied in patients on long-term statin therapy at the time of zoledronic acid infusion.¹⁸ The patient was also taking vitamin D₃ supplementation and was nearly in the reference range.⁵ His ED treatment included IV fluids and acetaminophen. Pretreatment (prior to or at the time of zoledronic acid infusion) with acetaminophen or ibuprofen may have prevented his symptoms, or at least lessened them to the point that an ED visit would not have resulted. The endocrinologist who prescribed the zoledronic acid documented a detailed discussion of the adverse effects of zoledronic acid with the patient, and the initial nursing call documents consideration of acute phase reaction. It is unclear whether the persistence of symptoms or worsening of symptoms ultimately led to the ED visit. Because no treatment was offered, it is unknown whether earlier posttreatment with acetaminophen, ibuprofen, or dexamethasone might have prevented his ED visit.

Conclusions

Clinicians who treat patients with osteoporosis should be aware of several key points. First, acute phase reaction symptoms are common with bisphosphonates, especially zoledronic acid infusions. Second, the symptoms are nonspecific but should have a suggestive time course. Third, dexamethasone may be partially protective, but based on the various trials discussed, it likely needs to be given for multiple days (instead of a single dose on the day of infusion). Given that acetaminophen and NSAIDs also seem to be protective (when given for multiple days starting on the day of infusion), both have lower overall adverse effect profiles than dexamethasone, consideration may be given to using either of these prophylactically.⁶ Dexamethasone could then be prescribed if symptoms are severe or persistent despite the use of acetaminophen or NSAIDs.

A 62-year-old man presented to the emergency department (ED) with 3 days of chills, myalgias, and nausea. The patient’s oral temperature at home ranged from 99.9 to 100.1 °F. He came to the ED after multiple phone discussions with primary care nursing over 3 days. His medical history included posttraumatic stress disorder, enlarged prostate, osteoporosis, gastroesophageal reflux, glaucoma, and left eye central retinal vein occlusion. Medications included fluoxetine 20 mg twice daily, omeprazole 20 mg twice daily, rosuvastatin 10 mg once daily, tamsulosin 0.4 mg nightly, and zolpidem 10 mg nightly. The patient’s glaucoma had been treated with a dexamethasone intraocular implant about 90 days earlier. The patient started on intravenous (IV) zoledronic acid for osteoporosis, with the first infusion 5 days prior to presentation.

In the ED, the patient’s temperature was 98.2 °F, blood pressure was 156/76 mm Hg, pulse was 94 bpm, respiratory rate was 16 breaths per minute, and 98% oxygen saturation on room air. He was in no acute distress, with an unremarkable physical examination reporting no abnormal respiratory sounds, no arrhythmia, normal gait, and no focal neurologic deficits. A comprehensive metabolic panel was unremarkable, creatine phosphokinase was 155 U/L (reference range, 30-240 U/L), and the complete blood count was notable only for an elevated white blood count of 15.3 × 10⁹/L (reference range, 4.0-11.0 × 10⁹/L), with 73.4% neutrophils, 16.2% lymphocytes, 9.1% monocytes, 0.5% eosinophils, and 0.4% basophils. The patient’s urinalysis was unremarkable.

What is your diagnosis? 

How would you treat this patient?

The ED physician considered viral infection and tested for both influenza and COVID-19. Laboratory results eliminated urinary tract infection and rhabdomyolysis as possible diagnoses. An acute phase reaction to zoledronic acid was determined to be the most likely cause. The patient was treated with IV saline in the ED, and acetaminophen both in the ED and at home.

Although initial nursing triage notes document consideration of acute phase reaction to zoledronic acid, the endocrinology service, which had recommended and arranged the zoledronic acid infusion, was not immediately notified of the reaction. It does not appear any treatment (eg, acetaminophen) was suggested, only that the patient was given advice this may resolve over 3 to 4 days. When he was seen 2 months later for an endocrinology follow-up appointment, he reported that all symptoms (chills, myalgias, and nausea) resolved gradually over 1 week. Since then, he has felt as well as he did before taking zoledronic acid. However, the patient was wary of further zoledronic acid, opting to defer deciding on a second dose until a future appointment.

Prior to starting zoledronic acid therapy, the patient was being treated for vitamin D deficiency. Four months prior to infusion, his 25-hydroxyvitamin D level was 12.0 ng/mL (reference range, 30 to 80 ng/mL). He then started taking cholecalciferol 100 mcg (4000 IU) daily. Eight days prior to infusion his 25-hydroxyvitamin D level was 29.5 ng/mL.

Federal health care practitioners, especially those working in the Veterans Health Administration (VHA), will commonly encounter patients similar to this case. Osteoporosisis is common in the United States with > 10 million diagnoses (including > 2 million men) and in VHA primary care populations.^1,2 Zoledronic acid is a frequently prescribed treatment, appearing in guidelines for osteoporosis management.^3-5

The acute phase reaction is a common adverse effect of both oral and IV bisphosphonates, although it’s substantially more common with IV bisphosphonates such as zoledronic acid. This reaction is characterized by flu-like symptoms of fever, myalgia, and arthralgia that occur within the first few days following bisphosphonate administration, and tends to be rated mild to moderate by patients.⁶ Clinical trial data from > 7000 women with postmenopausal osteoporosis found that 42% experienced ≥ 1 acute phase symptom following the first infusion (fever was most common, followed by musculoskeletal symptoms and gastrointestinal symptoms), compared with 12% for placebo. Incidence decreases with each subsequent infusion.⁷ Risk factors for reactions include low 25-hydroxyvitamin D levels,^8,9 no prior bisphosphonate exposure,⁹ younger age (aged 64-67 years vs 78-89 years),⁷ lower body mass index,¹⁰and higher lymphocyte levels at baseline.¹¹ While most cases are mild and self-limited, severe consequences have been noted, such as precipitation of adrenal crisis.^12,13 Additionally, more prolonged bone pain, sometimes quite severe, has been rarely reported with bisphosphonate use. However, it’s unclear whether this represents a separate adverse effect or a more severe acute phase reaction.⁶

The acute phase reaction is a transient inflammatory state marked by increases in proinflammatory cytokines such as C-reactive protein, interleukin-6, and tumor necrosis factor-α. Proposed mechanisms include: (1) inhibition of farnesyl pyrophosphate synthase, an enzyme of the mevalonate pathway, resulting inactivation of γϐ T cells and increased production of proinflammatory cytokines; (2) inhibition of the suppressor of cytokine signalling-3 in the macrophages, resulting in cessation of the suppression in cytokine signaling; or (3) negative regulation of γϐ T-cell expansion and interferon-c production by low serum 25-hydroxyvitamin D concentrations.¹¹

Prevention

Can an acute phase reaction to zoledronic acid be prevented? Bourke and colleagues reported that baseline calcium and/or vitamin D intake do not appear to affect rates of acute phase reaction in data pooled from 2 trials of zoledronic acid in postmenopausal women.¹⁴ However, patients receiving zoledronic acid had 25-hydroxyvitamin D values > 20 ng/mL 86% of the time, and values > 30 ng/mL 36% of the time. Bourke and colleagues suggest that “coadministration of calcium and vitamin D with zoledronate may not be necessary for individuals not at risk of marked vitamin D deficiency.”¹⁴ However, they did not prospectively test this hypothesis.

In our patient, vitamin D deficiency had been identified and treated, nearly achieving 30 ng/mL. The 2020 guidelines for postmenopausal osteoporosis recommend maintaining serum 25-hydroxyvitamin D levels 30 to 50 ng/mL, advising to supplement with vitamin D₃ as needed.⁵ The 2012 guidelines for osteoporosis in men from the Endocrine Society suggest that men with low vitamin D levels receive vitamin D supplements to raise the level > 30 ng/ml.⁴

Oral analgesics have been studied for the prevention of adverse effects related to zoledronic acid. Initiating 650 mg acetaminophen 45 minutes before zoledronic acid infusion and then every 6 hours over the next 3 days has been shown to significantly reduce symptoms.¹⁵ Acetaminophen or ibuprofen given every 6 hours for 3 days (starting 4 hours after zoledronic acid infusion) has been shown to reduce fever and other symptoms.¹⁶

Statins have been shown in vitro to prevent bisphosphonate-induced γϐ T cell activation.¹⁷ This has led to studies with various statins, although none have yet shown benefit in vivo. A double-blind, randomized, placebo-controlled trial of postmenopausal women for fluvastatin (single dose of 40 mg or 3 doses of 40 mg, each 24 hours apart) did not prevent acute phase reaction symptoms, nor did it prevent zoledronic acid-induced cytokine release.¹⁷ Rosuvastatin 10 mg daily starting 5 days before zoledronic acid treatment and taken for a total of 11 days did not show any difference in fever or pain.¹⁸ A protocol for pravastatin has been disseminated, but no study results have been published yet.¹⁹

Prophylactic dexamethasone has also been studied. A randomized double-blind, placebo-controlled trial of oral dexamethasone 4 mg at the time of first infusion of zoledronic acid found no significant difference in temperature change or symptom score over the following 3 days.²⁰ Chen and colleagues compared the efficacy of acetaminophen alone vs acetaminophen plus dexamethasone over several days.²¹ Acetaminophen 500 mg was given on the day of infusion and 4 times daily for 3 to 7 days for both groups, while dexamethasone 4 mg was given for 3 to 7 days. The dexamethasone group reported substantially lower incidence of any acute phase reaction symptoms (34% vs 67%, P = .003). A more recent study by Murdoch and colleagues comparing dexamethasone (4 mg daily for 3 days with the first dose 90 minutes before zoledronic acid infusion) with placebo found that the dexamethasone group had a statistically significant lower mean temperature change and acute phase reaction symptom score.²²

Adverse Effect Treatment

Treatment after development of acute phase reaction due to zoledronic acid infusion is generally limited to supportive care and/or nonsteroidal anti-inflammatory drugs (NSAIDs) acetaminophen or dexamethasone, largely based on extrapolation of the noted preventive trials and expert opinion.^3,6 Experiencing an acute phase reaction may portend better fracture risk reduction from zoledronic acid, although there is a potential association between acute phase reaction and mortality risk.^23,24

Our case was typical for acute phase reaction to zoledronic acid. The patient was already taking rosuvastatin 10 mg daily for hypercholesterolemia as prescribed by his primary care physician. Rosuvastatin was not shown to prevent symptoms, although it was not studied in patients on long-term statin therapy at the time of zoledronic acid infusion.¹⁸ The patient was also taking vitamin D₃ supplementation and was nearly in the reference range.⁵ His ED treatment included IV fluids and acetaminophen. Pretreatment (prior to or at the time of zoledronic acid infusion) with acetaminophen or ibuprofen may have prevented his symptoms, or at least lessened them to the point that an ED visit would not have resulted. The endocrinologist who prescribed the zoledronic acid documented a detailed discussion of the adverse effects of zoledronic acid with the patient, and the initial nursing call documents consideration of acute phase reaction. It is unclear whether the persistence of symptoms or worsening of symptoms ultimately led to the ED visit. Because no treatment was offered, it is unknown whether earlier posttreatment with acetaminophen, ibuprofen, or dexamethasone might have prevented his ED visit.

Conclusions

Clinicians who treat patients with osteoporosis should be aware of several key points. First, acute phase reaction symptoms are common with bisphosphonates, especially zoledronic acid infusions. Second, the symptoms are nonspecific but should have a suggestive time course. Third, dexamethasone may be partially protective, but based on the various trials discussed, it likely needs to be given for multiple days (instead of a single dose on the day of infusion). Given that acetaminophen and NSAIDs also seem to be protective (when given for multiple days starting on the day of infusion), both have lower overall adverse effect profiles than dexamethasone, consideration may be given to using either of these prophylactically.⁶ Dexamethasone could then be prescribed if symptoms are severe or persistent despite the use of acetaminophen or NSAIDs.

Nurses are recognized among the most trusted professions in the United States.¹ Since the time of Florence Nightingale, nurses have been challenged to provide care to patients and soldiers with complex needs, including acute and chronic physical illness, as well as mental health issues. Nurses have traditionally met those challenges with perseverance and creativity but have also experienced stress and burnout.

A shortage of nurses has been linked to many interrelated factors including the retirement of bedside caregivers and educators, diverse care settings, expanding roles for nurses, and nurse burnout.^2-4 Therefore, there is a critical need to better understand of how nurses can be supported while they care for patients, cope with stress, and maintain positive personal and professional outcomes. The objective of this pilot project was to assess US Department of Veterans Affairs (VA) nurses’ levels of burnout and test an intervention to enhance resourcefulness skills during the COVID-19 pandemic.

Background

Stress has many definitions. Hans Selye described it as a biological response of the body to any demand.^5,6 Occupational stress is a process that occurs in which work environment stressors result in the development of psychological, behavioral, or physiological effects that can contribute to health.⁶ Occupational stress has been observed as prevalent among nurses.⁶ In 1960, Menzies identified sources of stress among nurses that include complex decision-making within a dynamic environment.⁷ Since the mid-1980s, nurses’ stress at work has increased because of legal, accreditation, ethical issues, fiscal pressures, staffing shortages, and the increasing integration of technology associated with clinical care.⁸

Sustained stress can lead to emotional exhaustion or burnout, which has been associated with nursing turnover, lower patient satisfaction, and patient safety risk.^2,9 An American Nurses Foundation survey reported that 51% of US nurses feel exhausted, 43% overwhelmed, and 36% anxious; 28% express willingness to leave the profession.² Burnout has been described as a response to physical or emotional stress leading to exhaustion, self-doubt, cynicism, and ineffectiveness.¹⁰ Employees with burnout are more likely to leave their jobs, take sick leave, and suffer from depression and relationship problems,and it affects nearly half of all US nurses, especially among critical care, pediatric, and oncology specialities.^10,11 It has been well documented that unmitigated stress can lead to burnout and contribute to nurses leaving bedside care and the health care profession.^2,3 Several studies on nursing stress and burnout have focused on its prevalence and negative outcomes.^4,7,9 However, few studies have addressed building resiliency and resourcefulness for nurses.^10,12,13

A 2021 National Academy of Medicine report advocated a multilevel approach to managing burnout and building resiliency among nurses.¹⁴ Taylor further identified specific interventions, ranging from primary prevention to treatment.¹⁵ Primary prevention could include educating nurses on self-awareness, coping strategies, and communication skills. Screening for burnout and providing resources for support would be a secondary level of intervention. For nurses who experienced severe burnout symptoms and left the workplace, strategies are sorely needed to provide healing and a return-to-work plan.¹⁵ This may include adjusting nurse schedules and nursing roles (such as admitting/discharge nurse or resource nurse).

RESILIENCY AND RESOURCEFULNESS

Rushton and colleagues describe resiliency as the “ability to face adverse situations, remain focused, and continue to be optimistic for the future.”⁴ For nurses in complex health care systems, resiliency is associated with reduced turnover and symptoms of burnout and improved mental health. Humans are thought to have an innate resiliency potential that evolves over time and fluctuates depending on the context (eg, societal conditions, moral/ethical values, commitments).⁴ It is believed that resiliency can contribute to the development of new neuropathways that can be used to manage and cope with stress, prevent burnout, and improve quality of life. However, it appears these adaptations are individualized and contingent on situations, available resources, and changing priorities.¹⁶ Consequently, resiliency may be an essential tool for nurses to combat burnout in today’s complex health care systems.¹⁷

Although resilience and resourcefulness are conceptually related, each has distinctive features.¹⁸ Celinski frames resilience as transcendence and resourcefulness as transformation.¹⁹ Thus, while resilience suggests transcendence in terms of rising above, going beyond, exceeding, or excelling; resourcefulness reflects transformation, such as making changes in thoughts, feelings, behaviors, actions, or reactions. Resourcefulness has been conceptualized as an indicator of resilience.¹⁸

Resourcefulness comprises 2 dimensions, including the use of self-help strategies (personal resourcefulness) and seeking help from others (social resourcefulness), to self-regulate one’s thoughts, feelings, and behaviors to cope with high levels of stress, anxiety, or depression.^18,20,21 Personal resourcefulness skills include the use of cognitive reframing, positive thinking, problem-solving, priority-setting, and planning ahead. Social resourcefulness involves actively seeking help from others. Formal sources of help include, but are not limited to, nursing and medical care practitioners and community organizations such as hospitals and clinics. Informal sources of help include family members, friends, peers, and coworkers.

During the COVID-19 pandemic, nurses were especially challenged to provide support for each other because of limited nursing staff and treatment options, increased complex patient assignments, shortages of supplies, and reduced support services. Many nurses, however, were able to find innovative, peer-supported strategies for coping.¹³ Nurses’ use of resourcefulness skills is believed to be indicative of their resilience. This pilot project aimed to identify and evaluate some of these strategies and resourcefulness skills.

INTERVENTION

This pilot study among VA Northeast Ohio Health Care System (VANEOHS) nurses was designed to assess nursing burnout and resourcefulness during the pandemic. Those who agreed to participate completed a baseline survey on burnout and resourcefulness. Participants agreed to review a training video on resourcefulness skills (eg, relying on and exchanging ideas with others, and reframing and using ‘positive self-talk’). They were encouraged to document their experience with familiar and new resourcefulness skills. Weekly reminders (eg, emails and phone messages) reminded and coached participants in their journey.

The study identified and implemented an existing Resourcefulness Training (RT) intervention, which was developed for informal family caregivers and found to be effective.²² We measured burnout and resourcefulness preintervention and postintervention.²³ This survey and educational intervention were reviewed by the VANEOHS institutional review board and ruled exempt. The survey also gathered information on nurses' contact with individuals infected with COVID-19.

Despite the many staffing and resource challenges during the COVID-19 pandemic, a convenience sample of 12 nurses was recruited from nursing committees that continued to have scheduled meetings. These meetings allowed time to answer questions and provide information about the study. The majority of nurses queried declined to participate, citing no time, interest, or burnout. Participants completed a baseline survey, reviewed a 30-minute RT video, and tracked their experience for 28 days. Participants completed postintervention surveys 6 weeks after the video. Details of the survey and measures can be found in previous studies.^20,21
 

RT is an online cognitive-behavioral intervention that teaches and reinforces personal (self-help) and social (help-seeking) resourcefulness skills that have not yet been tested in nurses or other health care professionals.^22,24 The training included social resourcefulness (eg, from family, friends, others, and professionals) and personal resourcefulness (eg, problem-solving, positive thinking, self-control, organization skills). Participants were encouraged to review the videos as often as they preferred during these 4 weeks.

All 12 survey respondents were female and had received COVID-19 vaccinations according to the federal policy at the time of data collection. The number of patients cared for with COVID-19 infections varied widely (range, 1-1000). The baseline burnout score ranged from 1 (no burnout) to 3 (1 symptom of burnout, such as physical and emotional exhaustion), with a mean score of 2.2. In the follow-up survey, the mean score was 2.0. At baseline, participants reported a variety of activities to manage stress and burnout, including times with friends and family, engaging in hobbies, and prayer. Postintervention, some participants mentioned using skills learned from RT, including reframing the situation positively by refocusing and putting stressors in perspective (Table 1).

DISCUSSION

Recent American Nurses Association efforts to develop organizational and professional goals include the importance of nurses to recognize and manage stress to prevent burnout.²⁵ The American Nurses Association Code of Ethics notes that nurses have the same duties to care for themselves as they do for others.²⁵ Nurses have demonstrated the ability to adapt and remain resilient during stressful times. VA nurses are a resourceful group. Many used resourcefulness skills to manage stress and burnout even before the pandemic. For example, nurses identified using family/friends for support and validation, as well as prayer and meditation. Some of the new activities may have been influenced/inspired by RT, such as organizing schedules for problem-solving and distraction.

Relying on family and peers emerged as an essential resourcefulness skill. Support from peers—battle buddies—has been recognized as a key strategy among combat soldiers. A battle buddy is paired with a fellow soldier for support to keep each other informed about key instructions and information. This promotes cooperative problem-solving. Outcomes associated with battle buddies include increased morale and confidence, and decreased stress.²⁵ Over time, it is hoped that these coaching/mentoring relationships will result in enhanced leadership skills. Battle buddy strategies are currently being adapted into health care environments.^12,26 Such programs need to be further evaluated and information disseminated.

Findings from this pilot program support the use of interventions such as RT to decrease burnout among nurses. This study suggests that RT should be tested in a larger more representative sample to determine efficacy.

Limitations

This pilot study was limited by its small sample size, single facility, and female-only participants; the findings are not generalizable. Nurses were recruited from VA nursing committees and may not be representative of nurses in the general population. In addition, the RT intervention may require a longer time commitment to adequately determine efficacy. Another limitation was that personal or family exposure to COVID-19 was not measured, but may be a confounding variable. An additional limitation may have been the time interval. A baseline survey was completed prior to watching the teaching video. Daily logs were to be completed for 28 days. A post survey followed at 6 weeks. It is possible that there was insufficient time for the nurses to have the opportunity to use their resourcefulness skills within the short time frame of the study. While it supports the need for further studies, findings should be interpreted cautiously and not generalized. It may be premature based on these findings to conclude that the intervention will be effective for other populations. Further studies are needed to assess nurses’ preferences for healthy coping mechanisms, including RT.

Conclusions

As the nursing shortage continues, efforts to support diverse, innovative coping strategies remain a priority postpandemic. Nurses must be vigilant in appraising and managing their ability to cope and adapt to individual stress, while also being aware of the stress their colleagues are experiencing. Educational institutions, professional organizations, and health care facilities must strive to educate and support nurses to identify not only stress, but healthy coping mechanisms.

Acknowledgments

This work was supported by the US Department of Veterans Affairs Central Office rapid response COVID-19 funding initiative, the Veteran Affairs Northeast Ohio Health Care System, and Geriatric Research, Education, and Clinical Center (GRECC). The Resourcefulness Scale, Resourcefulness Skills Scale, and the Resourcefulness Training intervention are copyrighted and were used with permission of the copyright holder, Jaclene A. Zauszniewski, PhD, RN-BC.

Nurses are recognized among the most trusted professions in the United States.¹ Since the time of Florence Nightingale, nurses have been challenged to provide care to patients and soldiers with complex needs, including acute and chronic physical illness, as well as mental health issues. Nurses have traditionally met those challenges with perseverance and creativity but have also experienced stress and burnout.

A shortage of nurses has been linked to many interrelated factors including the retirement of bedside caregivers and educators, diverse care settings, expanding roles for nurses, and nurse burnout.^2-4 Therefore, there is a critical need to better understand of how nurses can be supported while they care for patients, cope with stress, and maintain positive personal and professional outcomes. The objective of this pilot project was to assess US Department of Veterans Affairs (VA) nurses’ levels of burnout and test an intervention to enhance resourcefulness skills during the COVID-19 pandemic.

Background

Stress has many definitions. Hans Selye described it as a biological response of the body to any demand.^5,6 Occupational stress is a process that occurs in which work environment stressors result in the development of psychological, behavioral, or physiological effects that can contribute to health.⁶ Occupational stress has been observed as prevalent among nurses.⁶ In 1960, Menzies identified sources of stress among nurses that include complex decision-making within a dynamic environment.⁷ Since the mid-1980s, nurses’ stress at work has increased because of legal, accreditation, ethical issues, fiscal pressures, staffing shortages, and the increasing integration of technology associated with clinical care.⁸

Sustained stress can lead to emotional exhaustion or burnout, which has been associated with nursing turnover, lower patient satisfaction, and patient safety risk.^2,9 An American Nurses Foundation survey reported that 51% of US nurses feel exhausted, 43% overwhelmed, and 36% anxious; 28% express willingness to leave the profession.² Burnout has been described as a response to physical or emotional stress leading to exhaustion, self-doubt, cynicism, and ineffectiveness.¹⁰ Employees with burnout are more likely to leave their jobs, take sick leave, and suffer from depression and relationship problems,and it affects nearly half of all US nurses, especially among critical care, pediatric, and oncology specialities.^10,11 It has been well documented that unmitigated stress can lead to burnout and contribute to nurses leaving bedside care and the health care profession.^2,3 Several studies on nursing stress and burnout have focused on its prevalence and negative outcomes.^4,7,9 However, few studies have addressed building resiliency and resourcefulness for nurses.^10,12,13

A 2021 National Academy of Medicine report advocated a multilevel approach to managing burnout and building resiliency among nurses.¹⁴ Taylor further identified specific interventions, ranging from primary prevention to treatment.¹⁵ Primary prevention could include educating nurses on self-awareness, coping strategies, and communication skills. Screening for burnout and providing resources for support would be a secondary level of intervention. For nurses who experienced severe burnout symptoms and left the workplace, strategies are sorely needed to provide healing and a return-to-work plan.¹⁵ This may include adjusting nurse schedules and nursing roles (such as admitting/discharge nurse or resource nurse).

RESILIENCY AND RESOURCEFULNESS

Rushton and colleagues describe resiliency as the “ability to face adverse situations, remain focused, and continue to be optimistic for the future.”⁴ For nurses in complex health care systems, resiliency is associated with reduced turnover and symptoms of burnout and improved mental health. Humans are thought to have an innate resiliency potential that evolves over time and fluctuates depending on the context (eg, societal conditions, moral/ethical values, commitments).⁴ It is believed that resiliency can contribute to the development of new neuropathways that can be used to manage and cope with stress, prevent burnout, and improve quality of life. However, it appears these adaptations are individualized and contingent on situations, available resources, and changing priorities.¹⁶ Consequently, resiliency may be an essential tool for nurses to combat burnout in today’s complex health care systems.¹⁷

Although resilience and resourcefulness are conceptually related, each has distinctive features.¹⁸ Celinski frames resilience as transcendence and resourcefulness as transformation.¹⁹ Thus, while resilience suggests transcendence in terms of rising above, going beyond, exceeding, or excelling; resourcefulness reflects transformation, such as making changes in thoughts, feelings, behaviors, actions, or reactions. Resourcefulness has been conceptualized as an indicator of resilience.¹⁸

Resourcefulness comprises 2 dimensions, including the use of self-help strategies (personal resourcefulness) and seeking help from others (social resourcefulness), to self-regulate one’s thoughts, feelings, and behaviors to cope with high levels of stress, anxiety, or depression.^18,20,21 Personal resourcefulness skills include the use of cognitive reframing, positive thinking, problem-solving, priority-setting, and planning ahead. Social resourcefulness involves actively seeking help from others. Formal sources of help include, but are not limited to, nursing and medical care practitioners and community organizations such as hospitals and clinics. Informal sources of help include family members, friends, peers, and coworkers.

During the COVID-19 pandemic, nurses were especially challenged to provide support for each other because of limited nursing staff and treatment options, increased complex patient assignments, shortages of supplies, and reduced support services. Many nurses, however, were able to find innovative, peer-supported strategies for coping.¹³ Nurses’ use of resourcefulness skills is believed to be indicative of their resilience. This pilot project aimed to identify and evaluate some of these strategies and resourcefulness skills.

INTERVENTION

This pilot study among VA Northeast Ohio Health Care System (VANEOHS) nurses was designed to assess nursing burnout and resourcefulness during the pandemic. Those who agreed to participate completed a baseline survey on burnout and resourcefulness. Participants agreed to review a training video on resourcefulness skills (eg, relying on and exchanging ideas with others, and reframing and using ‘positive self-talk’). They were encouraged to document their experience with familiar and new resourcefulness skills. Weekly reminders (eg, emails and phone messages) reminded and coached participants in their journey.

The study identified and implemented an existing Resourcefulness Training (RT) intervention, which was developed for informal family caregivers and found to be effective.²² We measured burnout and resourcefulness preintervention and postintervention.²³ This survey and educational intervention were reviewed by the VANEOHS institutional review board and ruled exempt. The survey also gathered information on nurses' contact with individuals infected with COVID-19.

Despite the many staffing and resource challenges during the COVID-19 pandemic, a convenience sample of 12 nurses was recruited from nursing committees that continued to have scheduled meetings. These meetings allowed time to answer questions and provide information about the study. The majority of nurses queried declined to participate, citing no time, interest, or burnout. Participants completed a baseline survey, reviewed a 30-minute RT video, and tracked their experience for 28 days. Participants completed postintervention surveys 6 weeks after the video. Details of the survey and measures can be found in previous studies.^20,21
 

RT is an online cognitive-behavioral intervention that teaches and reinforces personal (self-help) and social (help-seeking) resourcefulness skills that have not yet been tested in nurses or other health care professionals.^22,24 The training included social resourcefulness (eg, from family, friends, others, and professionals) and personal resourcefulness (eg, problem-solving, positive thinking, self-control, organization skills). Participants were encouraged to review the videos as often as they preferred during these 4 weeks.

All 12 survey respondents were female and had received COVID-19 vaccinations according to the federal policy at the time of data collection. The number of patients cared for with COVID-19 infections varied widely (range, 1-1000). The baseline burnout score ranged from 1 (no burnout) to 3 (1 symptom of burnout, such as physical and emotional exhaustion), with a mean score of 2.2. In the follow-up survey, the mean score was 2.0. At baseline, participants reported a variety of activities to manage stress and burnout, including times with friends and family, engaging in hobbies, and prayer. Postintervention, some participants mentioned using skills learned from RT, including reframing the situation positively by refocusing and putting stressors in perspective (Table 1).

DISCUSSION

Recent American Nurses Association efforts to develop organizational and professional goals include the importance of nurses to recognize and manage stress to prevent burnout.²⁵ The American Nurses Association Code of Ethics notes that nurses have the same duties to care for themselves as they do for others.²⁵ Nurses have demonstrated the ability to adapt and remain resilient during stressful times. VA nurses are a resourceful group. Many used resourcefulness skills to manage stress and burnout even before the pandemic. For example, nurses identified using family/friends for support and validation, as well as prayer and meditation. Some of the new activities may have been influenced/inspired by RT, such as organizing schedules for problem-solving and distraction.

Relying on family and peers emerged as an essential resourcefulness skill. Support from peers—battle buddies—has been recognized as a key strategy among combat soldiers. A battle buddy is paired with a fellow soldier for support to keep each other informed about key instructions and information. This promotes cooperative problem-solving. Outcomes associated with battle buddies include increased morale and confidence, and decreased stress.²⁵ Over time, it is hoped that these coaching/mentoring relationships will result in enhanced leadership skills. Battle buddy strategies are currently being adapted into health care environments.^12,26 Such programs need to be further evaluated and information disseminated.

Findings from this pilot program support the use of interventions such as RT to decrease burnout among nurses. This study suggests that RT should be tested in a larger more representative sample to determine efficacy.

Limitations

This pilot study was limited by its small sample size, single facility, and female-only participants; the findings are not generalizable. Nurses were recruited from VA nursing committees and may not be representative of nurses in the general population. In addition, the RT intervention may require a longer time commitment to adequately determine efficacy. Another limitation was that personal or family exposure to COVID-19 was not measured, but may be a confounding variable. An additional limitation may have been the time interval. A baseline survey was completed prior to watching the teaching video. Daily logs were to be completed for 28 days. A post survey followed at 6 weeks. It is possible that there was insufficient time for the nurses to have the opportunity to use their resourcefulness skills within the short time frame of the study. While it supports the need for further studies, findings should be interpreted cautiously and not generalized. It may be premature based on these findings to conclude that the intervention will be effective for other populations. Further studies are needed to assess nurses’ preferences for healthy coping mechanisms, including RT.

Conclusions

As the nursing shortage continues, efforts to support diverse, innovative coping strategies remain a priority postpandemic. Nurses must be vigilant in appraising and managing their ability to cope and adapt to individual stress, while also being aware of the stress their colleagues are experiencing. Educational institutions, professional organizations, and health care facilities must strive to educate and support nurses to identify not only stress, but healthy coping mechanisms.

Acknowledgments

This work was supported by the US Department of Veterans Affairs Central Office rapid response COVID-19 funding initiative, the Veteran Affairs Northeast Ohio Health Care System, and Geriatric Research, Education, and Clinical Center (GRECC). The Resourcefulness Scale, Resourcefulness Skills Scale, and the Resourcefulness Training intervention are copyrighted and were used with permission of the copyright holder, Jaclene A. Zauszniewski, PhD, RN-BC.

Nurses are recognized among the most trusted professions in the United States.¹ Since the time of Florence Nightingale, nurses have been challenged to provide care to patients and soldiers with complex needs, including acute and chronic physical illness, as well as mental health issues. Nurses have traditionally met those challenges with perseverance and creativity but have also experienced stress and burnout.

A shortage of nurses has been linked to many interrelated factors including the retirement of bedside caregivers and educators, diverse care settings, expanding roles for nurses, and nurse burnout.^2-4 Therefore, there is a critical need to better understand of how nurses can be supported while they care for patients, cope with stress, and maintain positive personal and professional outcomes. The objective of this pilot project was to assess US Department of Veterans Affairs (VA) nurses’ levels of burnout and test an intervention to enhance resourcefulness skills during the COVID-19 pandemic.

Background

Stress has many definitions. Hans Selye described it as a biological response of the body to any demand.^5,6 Occupational stress is a process that occurs in which work environment stressors result in the development of psychological, behavioral, or physiological effects that can contribute to health.⁶ Occupational stress has been observed as prevalent among nurses.⁶ In 1960, Menzies identified sources of stress among nurses that include complex decision-making within a dynamic environment.⁷ Since the mid-1980s, nurses’ stress at work has increased because of legal, accreditation, ethical issues, fiscal pressures, staffing shortages, and the increasing integration of technology associated with clinical care.⁸

Sustained stress can lead to emotional exhaustion or burnout, which has been associated with nursing turnover, lower patient satisfaction, and patient safety risk.^2,9 An American Nurses Foundation survey reported that 51% of US nurses feel exhausted, 43% overwhelmed, and 36% anxious; 28% express willingness to leave the profession.² Burnout has been described as a response to physical or emotional stress leading to exhaustion, self-doubt, cynicism, and ineffectiveness.¹⁰ Employees with burnout are more likely to leave their jobs, take sick leave, and suffer from depression and relationship problems,and it affects nearly half of all US nurses, especially among critical care, pediatric, and oncology specialities.^10,11 It has been well documented that unmitigated stress can lead to burnout and contribute to nurses leaving bedside care and the health care profession.^2,3 Several studies on nursing stress and burnout have focused on its prevalence and negative outcomes.^4,7,9 However, few studies have addressed building resiliency and resourcefulness for nurses.^10,12,13

A 2021 National Academy of Medicine report advocated a multilevel approach to managing burnout and building resiliency among nurses.¹⁴ Taylor further identified specific interventions, ranging from primary prevention to treatment.¹⁵ Primary prevention could include educating nurses on self-awareness, coping strategies, and communication skills. Screening for burnout and providing resources for support would be a secondary level of intervention. For nurses who experienced severe burnout symptoms and left the workplace, strategies are sorely needed to provide healing and a return-to-work plan.¹⁵ This may include adjusting nurse schedules and nursing roles (such as admitting/discharge nurse or resource nurse).

RESILIENCY AND RESOURCEFULNESS

Rushton and colleagues describe resiliency as the “ability to face adverse situations, remain focused, and continue to be optimistic for the future.”⁴ For nurses in complex health care systems, resiliency is associated with reduced turnover and symptoms of burnout and improved mental health. Humans are thought to have an innate resiliency potential that evolves over time and fluctuates depending on the context (eg, societal conditions, moral/ethical values, commitments).⁴ It is believed that resiliency can contribute to the development of new neuropathways that can be used to manage and cope with stress, prevent burnout, and improve quality of life. However, it appears these adaptations are individualized and contingent on situations, available resources, and changing priorities.¹⁶ Consequently, resiliency may be an essential tool for nurses to combat burnout in today’s complex health care systems.¹⁷

Although resilience and resourcefulness are conceptually related, each has distinctive features.¹⁸ Celinski frames resilience as transcendence and resourcefulness as transformation.¹⁹ Thus, while resilience suggests transcendence in terms of rising above, going beyond, exceeding, or excelling; resourcefulness reflects transformation, such as making changes in thoughts, feelings, behaviors, actions, or reactions. Resourcefulness has been conceptualized as an indicator of resilience.¹⁸

Resourcefulness comprises 2 dimensions, including the use of self-help strategies (personal resourcefulness) and seeking help from others (social resourcefulness), to self-regulate one’s thoughts, feelings, and behaviors to cope with high levels of stress, anxiety, or depression.^18,20,21 Personal resourcefulness skills include the use of cognitive reframing, positive thinking, problem-solving, priority-setting, and planning ahead. Social resourcefulness involves actively seeking help from others. Formal sources of help include, but are not limited to, nursing and medical care practitioners and community organizations such as hospitals and clinics. Informal sources of help include family members, friends, peers, and coworkers.

During the COVID-19 pandemic, nurses were especially challenged to provide support for each other because of limited nursing staff and treatment options, increased complex patient assignments, shortages of supplies, and reduced support services. Many nurses, however, were able to find innovative, peer-supported strategies for coping.¹³ Nurses’ use of resourcefulness skills is believed to be indicative of their resilience. This pilot project aimed to identify and evaluate some of these strategies and resourcefulness skills.

INTERVENTION

This pilot study among VA Northeast Ohio Health Care System (VANEOHS) nurses was designed to assess nursing burnout and resourcefulness during the pandemic. Those who agreed to participate completed a baseline survey on burnout and resourcefulness. Participants agreed to review a training video on resourcefulness skills (eg, relying on and exchanging ideas with others, and reframing and using ‘positive self-talk’). They were encouraged to document their experience with familiar and new resourcefulness skills. Weekly reminders (eg, emails and phone messages) reminded and coached participants in their journey.

The study identified and implemented an existing Resourcefulness Training (RT) intervention, which was developed for informal family caregivers and found to be effective.²² We measured burnout and resourcefulness preintervention and postintervention.²³ This survey and educational intervention were reviewed by the VANEOHS institutional review board and ruled exempt. The survey also gathered information on nurses' contact with individuals infected with COVID-19.

Despite the many staffing and resource challenges during the COVID-19 pandemic, a convenience sample of 12 nurses was recruited from nursing committees that continued to have scheduled meetings. These meetings allowed time to answer questions and provide information about the study. The majority of nurses queried declined to participate, citing no time, interest, or burnout. Participants completed a baseline survey, reviewed a 30-minute RT video, and tracked their experience for 28 days. Participants completed postintervention surveys 6 weeks after the video. Details of the survey and measures can be found in previous studies.^20,21
 

RT is an online cognitive-behavioral intervention that teaches and reinforces personal (self-help) and social (help-seeking) resourcefulness skills that have not yet been tested in nurses or other health care professionals.^22,24 The training included social resourcefulness (eg, from family, friends, others, and professionals) and personal resourcefulness (eg, problem-solving, positive thinking, self-control, organization skills). Participants were encouraged to review the videos as often as they preferred during these 4 weeks.

All 12 survey respondents were female and had received COVID-19 vaccinations according to the federal policy at the time of data collection. The number of patients cared for with COVID-19 infections varied widely (range, 1-1000). The baseline burnout score ranged from 1 (no burnout) to 3 (1 symptom of burnout, such as physical and emotional exhaustion), with a mean score of 2.2. In the follow-up survey, the mean score was 2.0. At baseline, participants reported a variety of activities to manage stress and burnout, including times with friends and family, engaging in hobbies, and prayer. Postintervention, some participants mentioned using skills learned from RT, including reframing the situation positively by refocusing and putting stressors in perspective (Table 1).

DISCUSSION

Recent American Nurses Association efforts to develop organizational and professional goals include the importance of nurses to recognize and manage stress to prevent burnout.²⁵ The American Nurses Association Code of Ethics notes that nurses have the same duties to care for themselves as they do for others.²⁵ Nurses have demonstrated the ability to adapt and remain resilient during stressful times. VA nurses are a resourceful group. Many used resourcefulness skills to manage stress and burnout even before the pandemic. For example, nurses identified using family/friends for support and validation, as well as prayer and meditation. Some of the new activities may have been influenced/inspired by RT, such as organizing schedules for problem-solving and distraction.

Relying on family and peers emerged as an essential resourcefulness skill. Support from peers—battle buddies—has been recognized as a key strategy among combat soldiers. A battle buddy is paired with a fellow soldier for support to keep each other informed about key instructions and information. This promotes cooperative problem-solving. Outcomes associated with battle buddies include increased morale and confidence, and decreased stress.²⁵ Over time, it is hoped that these coaching/mentoring relationships will result in enhanced leadership skills. Battle buddy strategies are currently being adapted into health care environments.^12,26 Such programs need to be further evaluated and information disseminated.

Findings from this pilot program support the use of interventions such as RT to decrease burnout among nurses. This study suggests that RT should be tested in a larger more representative sample to determine efficacy.

Limitations

This pilot study was limited by its small sample size, single facility, and female-only participants; the findings are not generalizable. Nurses were recruited from VA nursing committees and may not be representative of nurses in the general population. In addition, the RT intervention may require a longer time commitment to adequately determine efficacy. Another limitation was that personal or family exposure to COVID-19 was not measured, but may be a confounding variable. An additional limitation may have been the time interval. A baseline survey was completed prior to watching the teaching video. Daily logs were to be completed for 28 days. A post survey followed at 6 weeks. It is possible that there was insufficient time for the nurses to have the opportunity to use their resourcefulness skills within the short time frame of the study. While it supports the need for further studies, findings should be interpreted cautiously and not generalized. It may be premature based on these findings to conclude that the intervention will be effective for other populations. Further studies are needed to assess nurses’ preferences for healthy coping mechanisms, including RT.

Conclusions

As the nursing shortage continues, efforts to support diverse, innovative coping strategies remain a priority postpandemic. Nurses must be vigilant in appraising and managing their ability to cope and adapt to individual stress, while also being aware of the stress their colleagues are experiencing. Educational institutions, professional organizations, and health care facilities must strive to educate and support nurses to identify not only stress, but healthy coping mechanisms.

Acknowledgments

This work was supported by the US Department of Veterans Affairs Central Office rapid response COVID-19 funding initiative, the Veteran Affairs Northeast Ohio Health Care System, and Geriatric Research, Education, and Clinical Center (GRECC). The Resourcefulness Scale, Resourcefulness Skills Scale, and the Resourcefulness Training intervention are copyrighted and were used with permission of the copyright holder, Jaclene A. Zauszniewski, PhD, RN-BC.

Critical illness is commonly associated with interrelated conditions including pain, agitation, delirium, immobility, and sleep disruption (PADIS). Managing PADIS is often complex and includes pharmacologic and nonpharmacologic interventions.¹ Incorporating multifaceted practices to enhance PADIS management has been shown to improve several intensive care unit (ICU)-related outcomes.²

Many pharmacologic PADIS treatments are ineffective or associated with adverse effects. For example, antipsychotics used for treating ICU-related delirium have not shown improved outcomes.^3,4 Commonly used medications for agitation, such as benzodiazepines, increase delirium risk.^5,6 Because of these limitations, several nonpharmacologic interventions for PADIS have been evaluated.

Pet therapy has been implemented in some ICU settings, but is not widely adopted.⁷ Also referred to as animal-assisted activities, animal-assisted therapy, or animal-assisted interventions, pet therapy typically involves interaction between a patient and a live animal (most commonly a dog) under the direction of an animal handler, with the intention of providing therapeutic benefit. Interactions frequently include meet and greet activities such as petting, but also could include walking or other activities. Pet therapy has been reported to reduce pain, agitation, and stress among ICU patients.⁸ Introducing a pet therapy program with live animals in the ICU could be challenging because of factors such as identifying trained, accredited animals and handlers, and managing infection control and other risks.⁹ As an alternative to live pets, robotic pet therapy has been shown to be beneficial—mostly outside the ICU—in settings such as long-term care.^10,11 Although uncommon, robotic pets have been used in the ICU and hospital settings for therapeutic purposes.¹² Robotic pets reduce many concerns associated with live animals while mimicking the behaviors of live animals and potentially offering many of the same benefits.

OBSERVATIONS

The North Florida/South Georgia Veterans Health System (NF/SGVHS) implemented a novel robotic pet therapy program for patients requiring ICU care to improve the treatment of PADIS. Funding was provided through a Veterans Health Administration Innovation Grant procured by a clinical pharmacy specialist as the program’s champion. Goals of the robotic pet therapy program include reductions in: distressing symptoms associated with PADIS, use of psychoactive drugs and physical restraints, and ICU length of stay. The ICU team developed standard operating procedures and an order menu, which were integrated into the ICU prescriber ordering menu. Patients were selected for pet therapy based on PADIS scores and potential for positive response to pet therapy as assessed by the ICU team.Patients in medical and surgical ICU settings were eligible for the program. The robotic pets used in the program were Joy for AllCompanion Pets (Ageless Innovation LLC). Robotic cats and dogs were available and pets were “adopted’ by each patient (Figure). As an infection control measure, pets were not reissued or shared amongpatients and pets could be cleaned with a disinfectant solution. Nurses were primarily responsible for monitoring and documenting responses to robotic pet therapy.

It was necessary to secure buy-in from several services to successfully implement the program. The critical care clinical pharmacy specialists were responsible for ordering, storing, and dispensing the robotic pets. The NF/SGVHS innovation specialist helped secure funding, procure the robotic pet, and promote the program. The standard operating procedures for the program were developed by a multidisciplinary team with input from critical care nurses, intensivists, pharmacists, patient safety, and infection control (Table 1). Success of the program also required buy-in from ICU team members.

A retrospective cohort study was conducted to assess for improvements in PADIS symptoms and medication use post-intervention. Patients were included if they received robotic pet therapy in the ICU from July 10, 2019, to February 1, 2021. Individuals aged < 18 years or > 89 years, were pregnant, or were not receiving ICU-level care were excluded. Outcomes assessed included improvement in pain scores, agitation scores, sleep quality, resolution of delirium, and use of pain or psychoactive medications during patients’ ICU stay.

Thirty patients were included in the study (Table 2). After receiving a robotic pet, 9 (30%) patients recorded decreased pain scores, 15 (50%) recorded decreased agitation scores, 8 (27%) had resolution of delirium, and 2 (7%) described improvement in sleep. Pain medication use decreased in 12 (40%) patients and psychoactive medication use was reduced in 7 (23%) patients.

Limitations

The robotic pet therapy program has shown promising results; however, some aspects merit discussion. Evaluation of this program is limited by factors such as the observational study design, single-center patient sample, and lack of comparator group. Although no known adverse effects of robotic pet therapy were seen, it is possible that some patients may not have a favorable response. Challenges of implementing a robotic pet therapy program include cost and additional operational activities (storage, ordering, dispensing) necessary to maintain the program. Additional research is needed to evaluate the impact of robotic pet therapy on other outcomes including cost, ICU length of stay, and patient satisfaction.

CONCLUSIONS

Robotic pet therapy can be successfully implemented in the ICU and appears to provide a simple, safe, beneficial, nonpharmacologic intervention for PADIS. This study showed that many patients had favorable response to robotic pet therapy, indicating that it may be a viable alternative to traditional pet therapy. Other health systems could benefit from implementing programs similar to the robotic pet therapy program at NF/SGVHS.

Acknowledgments

The author would like to acknowledge Simran Panesar, PharmD, and Theresa Faison, PharmD, for their contributions to this project.

Critical illness is commonly associated with interrelated conditions including pain, agitation, delirium, immobility, and sleep disruption (PADIS). Managing PADIS is often complex and includes pharmacologic and nonpharmacologic interventions.¹ Incorporating multifaceted practices to enhance PADIS management has been shown to improve several intensive care unit (ICU)-related outcomes.²

Many pharmacologic PADIS treatments are ineffective or associated with adverse effects. For example, antipsychotics used for treating ICU-related delirium have not shown improved outcomes.^3,4 Commonly used medications for agitation, such as benzodiazepines, increase delirium risk.^5,6 Because of these limitations, several nonpharmacologic interventions for PADIS have been evaluated.

Pet therapy has been implemented in some ICU settings, but is not widely adopted.⁷ Also referred to as animal-assisted activities, animal-assisted therapy, or animal-assisted interventions, pet therapy typically involves interaction between a patient and a live animal (most commonly a dog) under the direction of an animal handler, with the intention of providing therapeutic benefit. Interactions frequently include meet and greet activities such as petting, but also could include walking or other activities. Pet therapy has been reported to reduce pain, agitation, and stress among ICU patients.⁸ Introducing a pet therapy program with live animals in the ICU could be challenging because of factors such as identifying trained, accredited animals and handlers, and managing infection control and other risks.⁹ As an alternative to live pets, robotic pet therapy has been shown to be beneficial—mostly outside the ICU—in settings such as long-term care.^10,11 Although uncommon, robotic pets have been used in the ICU and hospital settings for therapeutic purposes.¹² Robotic pets reduce many concerns associated with live animals while mimicking the behaviors of live animals and potentially offering many of the same benefits.

OBSERVATIONS

The North Florida/South Georgia Veterans Health System (NF/SGVHS) implemented a novel robotic pet therapy program for patients requiring ICU care to improve the treatment of PADIS. Funding was provided through a Veterans Health Administration Innovation Grant procured by a clinical pharmacy specialist as the program’s champion. Goals of the robotic pet therapy program include reductions in: distressing symptoms associated with PADIS, use of psychoactive drugs and physical restraints, and ICU length of stay. The ICU team developed standard operating procedures and an order menu, which were integrated into the ICU prescriber ordering menu. Patients were selected for pet therapy based on PADIS scores and potential for positive response to pet therapy as assessed by the ICU team.Patients in medical and surgical ICU settings were eligible for the program. The robotic pets used in the program were Joy for AllCompanion Pets (Ageless Innovation LLC). Robotic cats and dogs were available and pets were “adopted’ by each patient (Figure). As an infection control measure, pets were not reissued or shared amongpatients and pets could be cleaned with a disinfectant solution. Nurses were primarily responsible for monitoring and documenting responses to robotic pet therapy.

It was necessary to secure buy-in from several services to successfully implement the program. The critical care clinical pharmacy specialists were responsible for ordering, storing, and dispensing the robotic pets. The NF/SGVHS innovation specialist helped secure funding, procure the robotic pet, and promote the program. The standard operating procedures for the program were developed by a multidisciplinary team with input from critical care nurses, intensivists, pharmacists, patient safety, and infection control (Table 1). Success of the program also required buy-in from ICU team members.

A retrospective cohort study was conducted to assess for improvements in PADIS symptoms and medication use post-intervention. Patients were included if they received robotic pet therapy in the ICU from July 10, 2019, to February 1, 2021. Individuals aged < 18 years or > 89 years, were pregnant, or were not receiving ICU-level care were excluded. Outcomes assessed included improvement in pain scores, agitation scores, sleep quality, resolution of delirium, and use of pain or psychoactive medications during patients’ ICU stay.

Thirty patients were included in the study (Table 2). After receiving a robotic pet, 9 (30%) patients recorded decreased pain scores, 15 (50%) recorded decreased agitation scores, 8 (27%) had resolution of delirium, and 2 (7%) described improvement in sleep. Pain medication use decreased in 12 (40%) patients and psychoactive medication use was reduced in 7 (23%) patients.

Limitations

The robotic pet therapy program has shown promising results; however, some aspects merit discussion. Evaluation of this program is limited by factors such as the observational study design, single-center patient sample, and lack of comparator group. Although no known adverse effects of robotic pet therapy were seen, it is possible that some patients may not have a favorable response. Challenges of implementing a robotic pet therapy program include cost and additional operational activities (storage, ordering, dispensing) necessary to maintain the program. Additional research is needed to evaluate the impact of robotic pet therapy on other outcomes including cost, ICU length of stay, and patient satisfaction.

CONCLUSIONS

Robotic pet therapy can be successfully implemented in the ICU and appears to provide a simple, safe, beneficial, nonpharmacologic intervention for PADIS. This study showed that many patients had favorable response to robotic pet therapy, indicating that it may be a viable alternative to traditional pet therapy. Other health systems could benefit from implementing programs similar to the robotic pet therapy program at NF/SGVHS.

Acknowledgments

The author would like to acknowledge Simran Panesar, PharmD, and Theresa Faison, PharmD, for their contributions to this project.

Critical illness is commonly associated with interrelated conditions including pain, agitation, delirium, immobility, and sleep disruption (PADIS). Managing PADIS is often complex and includes pharmacologic and nonpharmacologic interventions.¹ Incorporating multifaceted practices to enhance PADIS management has been shown to improve several intensive care unit (ICU)-related outcomes.²

Many pharmacologic PADIS treatments are ineffective or associated with adverse effects. For example, antipsychotics used for treating ICU-related delirium have not shown improved outcomes.^3,4 Commonly used medications for agitation, such as benzodiazepines, increase delirium risk.^5,6 Because of these limitations, several nonpharmacologic interventions for PADIS have been evaluated.

Pet therapy has been implemented in some ICU settings, but is not widely adopted.⁷ Also referred to as animal-assisted activities, animal-assisted therapy, or animal-assisted interventions, pet therapy typically involves interaction between a patient and a live animal (most commonly a dog) under the direction of an animal handler, with the intention of providing therapeutic benefit. Interactions frequently include meet and greet activities such as petting, but also could include walking or other activities. Pet therapy has been reported to reduce pain, agitation, and stress among ICU patients.⁸ Introducing a pet therapy program with live animals in the ICU could be challenging because of factors such as identifying trained, accredited animals and handlers, and managing infection control and other risks.⁹ As an alternative to live pets, robotic pet therapy has been shown to be beneficial—mostly outside the ICU—in settings such as long-term care.^10,11 Although uncommon, robotic pets have been used in the ICU and hospital settings for therapeutic purposes.¹² Robotic pets reduce many concerns associated with live animals while mimicking the behaviors of live animals and potentially offering many of the same benefits.

OBSERVATIONS

The North Florida/South Georgia Veterans Health System (NF/SGVHS) implemented a novel robotic pet therapy program for patients requiring ICU care to improve the treatment of PADIS. Funding was provided through a Veterans Health Administration Innovation Grant procured by a clinical pharmacy specialist as the program’s champion. Goals of the robotic pet therapy program include reductions in: distressing symptoms associated with PADIS, use of psychoactive drugs and physical restraints, and ICU length of stay. The ICU team developed standard operating procedures and an order menu, which were integrated into the ICU prescriber ordering menu. Patients were selected for pet therapy based on PADIS scores and potential for positive response to pet therapy as assessed by the ICU team.Patients in medical and surgical ICU settings were eligible for the program. The robotic pets used in the program were Joy for AllCompanion Pets (Ageless Innovation LLC). Robotic cats and dogs were available and pets were “adopted’ by each patient (Figure). As an infection control measure, pets were not reissued or shared amongpatients and pets could be cleaned with a disinfectant solution. Nurses were primarily responsible for monitoring and documenting responses to robotic pet therapy.

It was necessary to secure buy-in from several services to successfully implement the program. The critical care clinical pharmacy specialists were responsible for ordering, storing, and dispensing the robotic pets. The NF/SGVHS innovation specialist helped secure funding, procure the robotic pet, and promote the program. The standard operating procedures for the program were developed by a multidisciplinary team with input from critical care nurses, intensivists, pharmacists, patient safety, and infection control (Table 1). Success of the program also required buy-in from ICU team members.

A retrospective cohort study was conducted to assess for improvements in PADIS symptoms and medication use post-intervention. Patients were included if they received robotic pet therapy in the ICU from July 10, 2019, to February 1, 2021. Individuals aged < 18 years or > 89 years, were pregnant, or were not receiving ICU-level care were excluded. Outcomes assessed included improvement in pain scores, agitation scores, sleep quality, resolution of delirium, and use of pain or psychoactive medications during patients’ ICU stay.

Thirty patients were included in the study (Table 2). After receiving a robotic pet, 9 (30%) patients recorded decreased pain scores, 15 (50%) recorded decreased agitation scores, 8 (27%) had resolution of delirium, and 2 (7%) described improvement in sleep. Pain medication use decreased in 12 (40%) patients and psychoactive medication use was reduced in 7 (23%) patients.

Limitations

The robotic pet therapy program has shown promising results; however, some aspects merit discussion. Evaluation of this program is limited by factors such as the observational study design, single-center patient sample, and lack of comparator group. Although no known adverse effects of robotic pet therapy were seen, it is possible that some patients may not have a favorable response. Challenges of implementing a robotic pet therapy program include cost and additional operational activities (storage, ordering, dispensing) necessary to maintain the program. Additional research is needed to evaluate the impact of robotic pet therapy on other outcomes including cost, ICU length of stay, and patient satisfaction.

CONCLUSIONS

Robotic pet therapy can be successfully implemented in the ICU and appears to provide a simple, safe, beneficial, nonpharmacologic intervention for PADIS. This study showed that many patients had favorable response to robotic pet therapy, indicating that it may be a viable alternative to traditional pet therapy. Other health systems could benefit from implementing programs similar to the robotic pet therapy program at NF/SGVHS.

Acknowledgments

The author would like to acknowledge Simran Panesar, PharmD, and Theresa Faison, PharmD, for their contributions to this project.

Following deployment to operations Desert Shield and Desert Storm (Gulf War) in 1990 and 1991, many Gulf War veterans (GWVs) developed chronic, complex symptoms, including pain, dyscognition, and fatigue, with gastrointestinal, skin, and respiratory manifestations. This Gulf War Illness (GWI) is reported to affect about 30% of those deployed. More than 30 years later, there is no consensus as to the etiology of GWI, although some deployment-related exposures have been implicated.¹

Accepted research definitions for GWI include the Centers for Disease Control and Prevention and Kansas definitions.² The US Department of Veterans Affairs (VA) uses the terminology chronic multisymptom illness (CMI), which is an overarching diagnosis under which GWI falls. Although there is no consensus case definition for CMI, there is overlap with conditions such as fibromyalgia, myalgic encephalomyelitis/chronic fatigue syndrome, and irritable bowel syndrome; the VA considers these as qualifying clinical diagnoses.³ The pathophysiology of GWI is also unknown, though a frequently reported unifying feature is that of autonomic nervous system (ANS) dysfunction. Studies have demonstrated differences between veterans with GWI and those without GWI in both the reporting of symptoms attributable to ANS dysfunction and in physiologic evaluations of the ANS.^4-10

Small fiber neuropathy (SFN), a condition with damage to the A-δ and C small nerve fibers, has been proposed as a potential mechanism for the pain and ANS dysfunction experienced in GWI.^11-13 Symptoms of SFN are similar to those of GWI, with pain and ANS symptoms commonly reported.^14,15 There are multiple diagnostic criteria for SFN, the most commonly used requiring the presence of appropriate symptoms in the absence of large fiber neuropathy and a skin biopsy demonstrating reduced intraepidermal nerve fiber density.^16-19 Several conditions reportedly cause SFN, most notably diabetes/prediabetes. Autoimmune disease, vitamin B12 deficiency, monoclonal gammopathies, celiac disease, paraneoplastic syndromes, and sodium channel gene mutations may also contribute to SFN.²⁰ Hyperlipidemia has been identified as a contributor, although it has been variably reported.^21,22

Idiopathic neuropathies, SFN included, may be secondary to neurotoxicant exposures. Agents whose exposure or consumption have been associated with SFN include alcohol most prominently, but also the organic solvent n-hexane, heavy metals, and excess vitamin B₆.^20,23-25 Agents associated with large fiber neuropathy may also have relevance for SFN, as small fibers have been likened to the “canary in the coal mine” in that they may be more susceptible to neurotoxicants and are affected earlier in the disease process.²⁶ In this way, SFN may be the harbinger of large fiber neuropathy in some cases. Of specific relevance for GWVs, organophosphates and carbamates are known to produce a delayed onset large fiber neuropathy.^27-30 Exposure to petrochemical solvents has also been associated with large fiber neuropathies.^31,32

The War Related Illness and Injury Study Center (WRIISC) is a clinical, research, and education center established by Congress in 2001. Its primary focus is on military exposures and postdeployment health of veterans. It is located at 3 sites: East Orange, New Jersey; Washington, DC; and Palo Alto, California. The New Jersey WRIISC began a program to evaluate GWVs with characteristic symptoms for possible SFN with use of a skin biopsy.

We hypothesize that SFN may underly much of GWI symptomatology and may not be accounted for by the putative etiologies detailed in review of the medical literature. This retrospective review of clinical evaluations for SFN in GWVs who sought care at the New Jersey WRIISC explored and addressed the following questions: (1) how common is biopsy-confirmed SFN in veterans with GWI; (2) do veterans with GWI and SFN report more symptoms attributable to ANS dysfunction when compared with veterans with GWI and no SFN; and (3) can SFN in veterans with GWI and SFN be explained by conditions and substances commonly associated with SFN? Institutional review board approval and waiver of consent was obtained from the Veterans Affairs New Jersey Health Care Center for the study.

Methods

A retrospective chart review was conducted on veterans evaluated at the WRIISC from March 1, 2015, to January 31, 2019. Inclusion criteria were: deployment to operations Desert Shield and Desert Storm between August 2, 1990, and February 28, 1991, and skin biopsy conducted at the WRIISC. Skin biopsies were obtained at the discretion of an examining clinician based on clinical indications, including neuropathic pain, ANS symptoms, and/or a fibromyalgia/chronic pain–type presentation.

Electronic health record review explicitly abstracted GWI status, results of the skin biopsy, and ANS symptom burden as determined by the Composite Autonomic Symptom Scale 31 (COMPASS 31) completed at the time of the WRIISC evaluation. Determination of GWI was established as per the clinical opinion of the WRIISC lead clinician or environmental exposure clinician as evidenced by a diagnosis of fibromyalgia or chronic fatigue syndrome, or explicit statement of CMI/GWI in the patient assessment. A diagnosis of SFN was established if clinical signs were present and an intraepidermal nerve fiber density below the lower limits, as compared to normative data from the clinical diagnostic laboratory (Therapath Neuropathology), was documented.

COMPASS 31 assesses symptoms across 6 domains (orthostatic, vasomotor, secretomotor, gastrointestinal, bladder, andpupillomotor). Patients are asked about symptom frequency (rarely to almost always), severity (mild to severe), and improvement (much worse to completely gone). Individual domain scores and a total weighted score (0-100) have demonstrated good validity, reliability, and consistency in SFN.^33,34

In veterans with GWI and documented SFN, a health record review was performed to identify potential etiologies for SFN (Appendix).

Statistical Analysis

Microsoft Excel and IBM SPSS 12.0.1 for Windows were used for data collection and statistical analysis. Fisher exact test was used for comparing the prevalence of SFN in veterans with GWI vs without GWI. The independent samples t test was used for comparing COMPASS 31 scores for veterans with GWI by SFN status. α < .05 was used for determining statistical significance. For those GWVs documented with SFN and GWI, potential explanations were documented in total and by condition.

Results

From March 1, 2015, to January 31, 2019, 141 GWVs received a comprehensive in person clinical evaluation at the WRIISC and 51 veterans (36%) received a skin biopsy and were included in this retrospective observational study (Figure). The mean age was 48.6 years, and the majority were male and served in the US Army. Skin biopsies met clinical criteria for GWI for 42 (82%) and 24 of 42 (57%) were determined to have SFN. Four of 9 (44%) veterans without GWI had positive SFN biopsies, though this difference was not statistically significant (Table 1). Veterans with SFN but no GWI were not included in the further analysis.

Thirty-five veterans with GWI—18 with SFN and 17 without SFN—completed the COMPASS 31 (Table 2). COMPASS 31 data were not analyzed for veterans without GWI. Individual domain scores and the difference in COMPASS 31 scores for veterans with GWI and SFN vs GWI and no SFN (38.3 vs 37.8, respectively) were not statistically significant.

Sixteen of 24 veterans with GWI and SFN (67%) had ≥ 1 conditions that could potentially be responsible for SFN (Table 3), including 11 veterans (46%) with prediabetes/diabetes. Hyperlipidemia is only variably reported as a cause of SFN; when included, 19 of 24 (79%) SFN cases were accounted for. We could not identify a medical explanation for SFN in 5 of 24 veterans (21%) with GWI, which were deemed to be idiopathic.

Discussion

Biopsy-confirmed SFN was present in more than half of our sample of veterans with GWI, which is broadly consistent with what has been reported in the literature.^13,35-38 In this clinical observation study, SFN was similarly prevalent in veterans with and without GWI; although it should be noted that biopsies only were obtained when there was a strong clinical suspicion for SFN. Almost half of patients with GWI did not have SFN, so our study does not support SFN as the underlying explanation for all GWI. Although our data cannot provide clinical guidance as to when skin biopsy may be indicated in GWI, work done in fibromyalgia found symptoms of dysautonomia and paresthesias are more specific for SFN and may be useful to help guide medical decision making.³⁹

Veterans with GWI in our clinical sample reported a high burden of clinical symptoms conceivably attributable to ANS dysfunction. This symptom reporting is consistent with that seen in other GWI studies, as well as in other studies of SFN.^{4,5,7-9,14,15,34,38,40} Our clinical sample of veterans with GWI found no differences in the ANS symptom reporting between those with and without SFN. Therefore, our study cannot support SFN alone as accounting for ANS symptom burden in patients with GWI.

Two-thirds of biopsy-confirmed SFN in our clinical sample of veterans with GWI could potentially be explained by established medical conditions. As in other studies of SFN, prediabetes and diabetes represented a plurality (46%). Even after considering hyperlipidemia as a potential explanation, about 21% of SFN cases in veterans with GWI still were deemed idiopathic.

Evidence supports certain environmental agents as causal factors for GWI. Neurotoxicants reportedly related to GWI include pesticides (particularly organophosphates and carbamates), pyridostigmine bromide (used during the Gulf War as a prophylactic agent against the use of chemical weapons), and low levels of the nerve agent sarin from environmental contamination due to chemical weapons detonations.¹ Some of these agents have been implicated in neuropathy as well.^1,28-30 It is biologically plausible that deployment-related exposures could trigger SFN, though the traditional consensus has been that remote exposure to neurotoxic substances is unlikely to produce neuropathy that presents many years after the exposure.⁴¹ In the WRIISC clinical experience, however, veterans often report that their neuropathic symptoms predate the diagnosis of the associated medical conditions, sometimes by decades. It is conceivable that remote exposures may trigger the condition that is then potentiated by ongoing exposures, metabolic factors, and/or other medical conditions. These may perpetuate neuropathic symptoms and the illness experience of affected veterans. Our clinical observation study cannot clarify the extent to which this may be the case. Despite these findings and arguments, an environmental contribution to SFN cannot be discounted, and further research is needed to explore a potential relationship.

Limitations

This study’s conclusions are limited by its observational/retrospective design in a relatively small clinical sample of veterans evaluated at a tertiary referral center for postdeployment exposure-related health concerns. The WRIISC clinical sample is not representative of all GWVs or even of all veterans with GWI, as there is inherent selection bias as to who gets referred to and evaluated at the WRIISC. As with studies based on retrospective chart review, data are reliant on clinical documentation andaccuracy/consistency of the reviewer. Evaluation for SFN with skin biopsy is an invasive procedure and was performed when a high index of clinical suspicion for this condition existed, possibly representing confirmation bias. Therefore, the relatively high prevalence ofbiopsy-confirmed SFN seen in our clinical sample cannot be generalized to GWVs as a whole or even to veterans with GWI.

Assessment of autonomic dysfunction was based on COMPASS 31 symptom reporting by an small subset of the clinical cohort. Symptom reporting may not be reflective of true abnormality in ANS function. Physiologic tests of the ANS were not performed; such studies could more objectively establish whether ANS dysfunction is more prevalent in GWI veterans with SFN.

Evaluation for all potential etiologic/contributory conditions to SFN was not exhaustive. For example, sodium channel gene mutations have been documented to account for up to one-third of all cases of idiopathic SFN.⁴² For those cases in which no compelling etiology was identified, it is plausible that medical explanations for SFN may be found on further investigation.

Clinical assessments at the WRIISC were performed on GWVs ≥ 26 years after their deployment-related exposures. Other conditions/exposures may have occurred in the interim. What is not clear is whether the SFN predated the onset of any of these medical conditions or other putative contributors. This observational study is not able to tease out a temporal association to make a cause-and-effect assessment.

Conclusions

Retrospective analysis of clinical data of veterans evaluated at a specialized center for postdeployment health demonstrated that skin biopsy–confirmed SFN was prevalent, but not ubiquitous, in veterans with GWI. Symptom that may be attributed to ANS dysfunction in this clinical sample was consistent with literature on SFN and with GWI, but we could not definitively attribute ANS symptoms to SFN. Our study does not support the hypothesis that GWI symptoms are solely due to SFN, though it may still be relevant in a subset of veterans with GWI with strongly suggestive clinical features. We were able to identify a potential etiology for SFN in most veterans with GWI. Further investigations are recommended to explore any potential relationship between Gulf War exposures and SFN.

Following deployment to operations Desert Shield and Desert Storm (Gulf War) in 1990 and 1991, many Gulf War veterans (GWVs) developed chronic, complex symptoms, including pain, dyscognition, and fatigue, with gastrointestinal, skin, and respiratory manifestations. This Gulf War Illness (GWI) is reported to affect about 30% of those deployed. More than 30 years later, there is no consensus as to the etiology of GWI, although some deployment-related exposures have been implicated.¹

Accepted research definitions for GWI include the Centers for Disease Control and Prevention and Kansas definitions.² The US Department of Veterans Affairs (VA) uses the terminology chronic multisymptom illness (CMI), which is an overarching diagnosis under which GWI falls. Although there is no consensus case definition for CMI, there is overlap with conditions such as fibromyalgia, myalgic encephalomyelitis/chronic fatigue syndrome, and irritable bowel syndrome; the VA considers these as qualifying clinical diagnoses.³ The pathophysiology of GWI is also unknown, though a frequently reported unifying feature is that of autonomic nervous system (ANS) dysfunction. Studies have demonstrated differences between veterans with GWI and those without GWI in both the reporting of symptoms attributable to ANS dysfunction and in physiologic evaluations of the ANS.^4-10

Small fiber neuropathy (SFN), a condition with damage to the A-δ and C small nerve fibers, has been proposed as a potential mechanism for the pain and ANS dysfunction experienced in GWI.^11-13 Symptoms of SFN are similar to those of GWI, with pain and ANS symptoms commonly reported.^14,15 There are multiple diagnostic criteria for SFN, the most commonly used requiring the presence of appropriate symptoms in the absence of large fiber neuropathy and a skin biopsy demonstrating reduced intraepidermal nerve fiber density.^16-19 Several conditions reportedly cause SFN, most notably diabetes/prediabetes. Autoimmune disease, vitamin B12 deficiency, monoclonal gammopathies, celiac disease, paraneoplastic syndromes, and sodium channel gene mutations may also contribute to SFN.²⁰ Hyperlipidemia has been identified as a contributor, although it has been variably reported.^21,22

Idiopathic neuropathies, SFN included, may be secondary to neurotoxicant exposures. Agents whose exposure or consumption have been associated with SFN include alcohol most prominently, but also the organic solvent n-hexane, heavy metals, and excess vitamin B₆.^20,23-25 Agents associated with large fiber neuropathy may also have relevance for SFN, as small fibers have been likened to the “canary in the coal mine” in that they may be more susceptible to neurotoxicants and are affected earlier in the disease process.²⁶ In this way, SFN may be the harbinger of large fiber neuropathy in some cases. Of specific relevance for GWVs, organophosphates and carbamates are known to produce a delayed onset large fiber neuropathy.^27-30 Exposure to petrochemical solvents has also been associated with large fiber neuropathies.^31,32

The War Related Illness and Injury Study Center (WRIISC) is a clinical, research, and education center established by Congress in 2001. Its primary focus is on military exposures and postdeployment health of veterans. It is located at 3 sites: East Orange, New Jersey; Washington, DC; and Palo Alto, California. The New Jersey WRIISC began a program to evaluate GWVs with characteristic symptoms for possible SFN with use of a skin biopsy.

We hypothesize that SFN may underly much of GWI symptomatology and may not be accounted for by the putative etiologies detailed in review of the medical literature. This retrospective review of clinical evaluations for SFN in GWVs who sought care at the New Jersey WRIISC explored and addressed the following questions: (1) how common is biopsy-confirmed SFN in veterans with GWI; (2) do veterans with GWI and SFN report more symptoms attributable to ANS dysfunction when compared with veterans with GWI and no SFN; and (3) can SFN in veterans with GWI and SFN be explained by conditions and substances commonly associated with SFN? Institutional review board approval and waiver of consent was obtained from the Veterans Affairs New Jersey Health Care Center for the study.

Methods

A retrospective chart review was conducted on veterans evaluated at the WRIISC from March 1, 2015, to January 31, 2019. Inclusion criteria were: deployment to operations Desert Shield and Desert Storm between August 2, 1990, and February 28, 1991, and skin biopsy conducted at the WRIISC. Skin biopsies were obtained at the discretion of an examining clinician based on clinical indications, including neuropathic pain, ANS symptoms, and/or a fibromyalgia/chronic pain–type presentation.

Electronic health record review explicitly abstracted GWI status, results of the skin biopsy, and ANS symptom burden as determined by the Composite Autonomic Symptom Scale 31 (COMPASS 31) completed at the time of the WRIISC evaluation. Determination of GWI was established as per the clinical opinion of the WRIISC lead clinician or environmental exposure clinician as evidenced by a diagnosis of fibromyalgia or chronic fatigue syndrome, or explicit statement of CMI/GWI in the patient assessment. A diagnosis of SFN was established if clinical signs were present and an intraepidermal nerve fiber density below the lower limits, as compared to normative data from the clinical diagnostic laboratory (Therapath Neuropathology), was documented.

COMPASS 31 assesses symptoms across 6 domains (orthostatic, vasomotor, secretomotor, gastrointestinal, bladder, andpupillomotor). Patients are asked about symptom frequency (rarely to almost always), severity (mild to severe), and improvement (much worse to completely gone). Individual domain scores and a total weighted score (0-100) have demonstrated good validity, reliability, and consistency in SFN.^33,34

In veterans with GWI and documented SFN, a health record review was performed to identify potential etiologies for SFN (Appendix).

Statistical Analysis

Microsoft Excel and IBM SPSS 12.0.1 for Windows were used for data collection and statistical analysis. Fisher exact test was used for comparing the prevalence of SFN in veterans with GWI vs without GWI. The independent samples t test was used for comparing COMPASS 31 scores for veterans with GWI by SFN status. α < .05 was used for determining statistical significance. For those GWVs documented with SFN and GWI, potential explanations were documented in total and by condition.

Results

From March 1, 2015, to January 31, 2019, 141 GWVs received a comprehensive in person clinical evaluation at the WRIISC and 51 veterans (36%) received a skin biopsy and were included in this retrospective observational study (Figure). The mean age was 48.6 years, and the majority were male and served in the US Army. Skin biopsies met clinical criteria for GWI for 42 (82%) and 24 of 42 (57%) were determined to have SFN. Four of 9 (44%) veterans without GWI had positive SFN biopsies, though this difference was not statistically significant (Table 1). Veterans with SFN but no GWI were not included in the further analysis.

Thirty-five veterans with GWI—18 with SFN and 17 without SFN—completed the COMPASS 31 (Table 2). COMPASS 31 data were not analyzed for veterans without GWI. Individual domain scores and the difference in COMPASS 31 scores for veterans with GWI and SFN vs GWI and no SFN (38.3 vs 37.8, respectively) were not statistically significant.

Sixteen of 24 veterans with GWI and SFN (67%) had ≥ 1 conditions that could potentially be responsible for SFN (Table 3), including 11 veterans (46%) with prediabetes/diabetes. Hyperlipidemia is only variably reported as a cause of SFN; when included, 19 of 24 (79%) SFN cases were accounted for. We could not identify a medical explanation for SFN in 5 of 24 veterans (21%) with GWI, which were deemed to be idiopathic.

Discussion

Biopsy-confirmed SFN was present in more than half of our sample of veterans with GWI, which is broadly consistent with what has been reported in the literature.^13,35-38 In this clinical observation study, SFN was similarly prevalent in veterans with and without GWI; although it should be noted that biopsies only were obtained when there was a strong clinical suspicion for SFN. Almost half of patients with GWI did not have SFN, so our study does not support SFN as the underlying explanation for all GWI. Although our data cannot provide clinical guidance as to when skin biopsy may be indicated in GWI, work done in fibromyalgia found symptoms of dysautonomia and paresthesias are more specific for SFN and may be useful to help guide medical decision making.³⁹

Veterans with GWI in our clinical sample reported a high burden of clinical symptoms conceivably attributable to ANS dysfunction. This symptom reporting is consistent with that seen in other GWI studies, as well as in other studies of SFN.^{4,5,7-9,14,15,34,38,40} Our clinical sample of veterans with GWI found no differences in the ANS symptom reporting between those with and without SFN. Therefore, our study cannot support SFN alone as accounting for ANS symptom burden in patients with GWI.

Two-thirds of biopsy-confirmed SFN in our clinical sample of veterans with GWI could potentially be explained by established medical conditions. As in other studies of SFN, prediabetes and diabetes represented a plurality (46%). Even after considering hyperlipidemia as a potential explanation, about 21% of SFN cases in veterans with GWI still were deemed idiopathic.

Evidence supports certain environmental agents as causal factors for GWI. Neurotoxicants reportedly related to GWI include pesticides (particularly organophosphates and carbamates), pyridostigmine bromide (used during the Gulf War as a prophylactic agent against the use of chemical weapons), and low levels of the nerve agent sarin from environmental contamination due to chemical weapons detonations.¹ Some of these agents have been implicated in neuropathy as well.^1,28-30 It is biologically plausible that deployment-related exposures could trigger SFN, though the traditional consensus has been that remote exposure to neurotoxic substances is unlikely to produce neuropathy that presents many years after the exposure.⁴¹ In the WRIISC clinical experience, however, veterans often report that their neuropathic symptoms predate the diagnosis of the associated medical conditions, sometimes by decades. It is conceivable that remote exposures may trigger the condition that is then potentiated by ongoing exposures, metabolic factors, and/or other medical conditions. These may perpetuate neuropathic symptoms and the illness experience of affected veterans. Our clinical observation study cannot clarify the extent to which this may be the case. Despite these findings and arguments, an environmental contribution to SFN cannot be discounted, and further research is needed to explore a potential relationship.

Limitations

This study’s conclusions are limited by its observational/retrospective design in a relatively small clinical sample of veterans evaluated at a tertiary referral center for postdeployment exposure-related health concerns. The WRIISC clinical sample is not representative of all GWVs or even of all veterans with GWI, as there is inherent selection bias as to who gets referred to and evaluated at the WRIISC. As with studies based on retrospective chart review, data are reliant on clinical documentation andaccuracy/consistency of the reviewer. Evaluation for SFN with skin biopsy is an invasive procedure and was performed when a high index of clinical suspicion for this condition existed, possibly representing confirmation bias. Therefore, the relatively high prevalence ofbiopsy-confirmed SFN seen in our clinical sample cannot be generalized to GWVs as a whole or even to veterans with GWI.

Assessment of autonomic dysfunction was based on COMPASS 31 symptom reporting by an small subset of the clinical cohort. Symptom reporting may not be reflective of true abnormality in ANS function. Physiologic tests of the ANS were not performed; such studies could more objectively establish whether ANS dysfunction is more prevalent in GWI veterans with SFN.

Evaluation for all potential etiologic/contributory conditions to SFN was not exhaustive. For example, sodium channel gene mutations have been documented to account for up to one-third of all cases of idiopathic SFN.⁴² For those cases in which no compelling etiology was identified, it is plausible that medical explanations for SFN may be found on further investigation.

Clinical assessments at the WRIISC were performed on GWVs ≥ 26 years after their deployment-related exposures. Other conditions/exposures may have occurred in the interim. What is not clear is whether the SFN predated the onset of any of these medical conditions or other putative contributors. This observational study is not able to tease out a temporal association to make a cause-and-effect assessment.

Conclusions

Retrospective analysis of clinical data of veterans evaluated at a specialized center for postdeployment health demonstrated that skin biopsy–confirmed SFN was prevalent, but not ubiquitous, in veterans with GWI. Symptom that may be attributed to ANS dysfunction in this clinical sample was consistent with literature on SFN and with GWI, but we could not definitively attribute ANS symptoms to SFN. Our study does not support the hypothesis that GWI symptoms are solely due to SFN, though it may still be relevant in a subset of veterans with GWI with strongly suggestive clinical features. We were able to identify a potential etiology for SFN in most veterans with GWI. Further investigations are recommended to explore any potential relationship between Gulf War exposures and SFN.

Following deployment to operations Desert Shield and Desert Storm (Gulf War) in 1990 and 1991, many Gulf War veterans (GWVs) developed chronic, complex symptoms, including pain, dyscognition, and fatigue, with gastrointestinal, skin, and respiratory manifestations. This Gulf War Illness (GWI) is reported to affect about 30% of those deployed. More than 30 years later, there is no consensus as to the etiology of GWI, although some deployment-related exposures have been implicated.¹

Accepted research definitions for GWI include the Centers for Disease Control and Prevention and Kansas definitions.² The US Department of Veterans Affairs (VA) uses the terminology chronic multisymptom illness (CMI), which is an overarching diagnosis under which GWI falls. Although there is no consensus case definition for CMI, there is overlap with conditions such as fibromyalgia, myalgic encephalomyelitis/chronic fatigue syndrome, and irritable bowel syndrome; the VA considers these as qualifying clinical diagnoses.³ The pathophysiology of GWI is also unknown, though a frequently reported unifying feature is that of autonomic nervous system (ANS) dysfunction. Studies have demonstrated differences between veterans with GWI and those without GWI in both the reporting of symptoms attributable to ANS dysfunction and in physiologic evaluations of the ANS.^4-10

Small fiber neuropathy (SFN), a condition with damage to the A-δ and C small nerve fibers, has been proposed as a potential mechanism for the pain and ANS dysfunction experienced in GWI.^11-13 Symptoms of SFN are similar to those of GWI, with pain and ANS symptoms commonly reported.^14,15 There are multiple diagnostic criteria for SFN, the most commonly used requiring the presence of appropriate symptoms in the absence of large fiber neuropathy and a skin biopsy demonstrating reduced intraepidermal nerve fiber density.^16-19 Several conditions reportedly cause SFN, most notably diabetes/prediabetes. Autoimmune disease, vitamin B12 deficiency, monoclonal gammopathies, celiac disease, paraneoplastic syndromes, and sodium channel gene mutations may also contribute to SFN.²⁰ Hyperlipidemia has been identified as a contributor, although it has been variably reported.^21,22

Idiopathic neuropathies, SFN included, may be secondary to neurotoxicant exposures. Agents whose exposure or consumption have been associated with SFN include alcohol most prominently, but also the organic solvent n-hexane, heavy metals, and excess vitamin B₆.^20,23-25 Agents associated with large fiber neuropathy may also have relevance for SFN, as small fibers have been likened to the “canary in the coal mine” in that they may be more susceptible to neurotoxicants and are affected earlier in the disease process.²⁶ In this way, SFN may be the harbinger of large fiber neuropathy in some cases. Of specific relevance for GWVs, organophosphates and carbamates are known to produce a delayed onset large fiber neuropathy.^27-30 Exposure to petrochemical solvents has also been associated with large fiber neuropathies.^31,32

The War Related Illness and Injury Study Center (WRIISC) is a clinical, research, and education center established by Congress in 2001. Its primary focus is on military exposures and postdeployment health of veterans. It is located at 3 sites: East Orange, New Jersey; Washington, DC; and Palo Alto, California. The New Jersey WRIISC began a program to evaluate GWVs with characteristic symptoms for possible SFN with use of a skin biopsy.

We hypothesize that SFN may underly much of GWI symptomatology and may not be accounted for by the putative etiologies detailed in review of the medical literature. This retrospective review of clinical evaluations for SFN in GWVs who sought care at the New Jersey WRIISC explored and addressed the following questions: (1) how common is biopsy-confirmed SFN in veterans with GWI; (2) do veterans with GWI and SFN report more symptoms attributable to ANS dysfunction when compared with veterans with GWI and no SFN; and (3) can SFN in veterans with GWI and SFN be explained by conditions and substances commonly associated with SFN? Institutional review board approval and waiver of consent was obtained from the Veterans Affairs New Jersey Health Care Center for the study.

Methods

A retrospective chart review was conducted on veterans evaluated at the WRIISC from March 1, 2015, to January 31, 2019. Inclusion criteria were: deployment to operations Desert Shield and Desert Storm between August 2, 1990, and February 28, 1991, and skin biopsy conducted at the WRIISC. Skin biopsies were obtained at the discretion of an examining clinician based on clinical indications, including neuropathic pain, ANS symptoms, and/or a fibromyalgia/chronic pain–type presentation.

Electronic health record review explicitly abstracted GWI status, results of the skin biopsy, and ANS symptom burden as determined by the Composite Autonomic Symptom Scale 31 (COMPASS 31) completed at the time of the WRIISC evaluation. Determination of GWI was established as per the clinical opinion of the WRIISC lead clinician or environmental exposure clinician as evidenced by a diagnosis of fibromyalgia or chronic fatigue syndrome, or explicit statement of CMI/GWI in the patient assessment. A diagnosis of SFN was established if clinical signs were present and an intraepidermal nerve fiber density below the lower limits, as compared to normative data from the clinical diagnostic laboratory (Therapath Neuropathology), was documented.

COMPASS 31 assesses symptoms across 6 domains (orthostatic, vasomotor, secretomotor, gastrointestinal, bladder, andpupillomotor). Patients are asked about symptom frequency (rarely to almost always), severity (mild to severe), and improvement (much worse to completely gone). Individual domain scores and a total weighted score (0-100) have demonstrated good validity, reliability, and consistency in SFN.^33,34

In veterans with GWI and documented SFN, a health record review was performed to identify potential etiologies for SFN (Appendix).

Statistical Analysis

Microsoft Excel and IBM SPSS 12.0.1 for Windows were used for data collection and statistical analysis. Fisher exact test was used for comparing the prevalence of SFN in veterans with GWI vs without GWI. The independent samples t test was used for comparing COMPASS 31 scores for veterans with GWI by SFN status. α < .05 was used for determining statistical significance. For those GWVs documented with SFN and GWI, potential explanations were documented in total and by condition.

Results

From March 1, 2015, to January 31, 2019, 141 GWVs received a comprehensive in person clinical evaluation at the WRIISC and 51 veterans (36%) received a skin biopsy and were included in this retrospective observational study (Figure). The mean age was 48.6 years, and the majority were male and served in the US Army. Skin biopsies met clinical criteria for GWI for 42 (82%) and 24 of 42 (57%) were determined to have SFN. Four of 9 (44%) veterans without GWI had positive SFN biopsies, though this difference was not statistically significant (Table 1). Veterans with SFN but no GWI were not included in the further analysis.

Thirty-five veterans with GWI—18 with SFN and 17 without SFN—completed the COMPASS 31 (Table 2). COMPASS 31 data were not analyzed for veterans without GWI. Individual domain scores and the difference in COMPASS 31 scores for veterans with GWI and SFN vs GWI and no SFN (38.3 vs 37.8, respectively) were not statistically significant.

Sixteen of 24 veterans with GWI and SFN (67%) had ≥ 1 conditions that could potentially be responsible for SFN (Table 3), including 11 veterans (46%) with prediabetes/diabetes. Hyperlipidemia is only variably reported as a cause of SFN; when included, 19 of 24 (79%) SFN cases were accounted for. We could not identify a medical explanation for SFN in 5 of 24 veterans (21%) with GWI, which were deemed to be idiopathic.

Discussion

Biopsy-confirmed SFN was present in more than half of our sample of veterans with GWI, which is broadly consistent with what has been reported in the literature.^13,35-38 In this clinical observation study, SFN was similarly prevalent in veterans with and without GWI; although it should be noted that biopsies only were obtained when there was a strong clinical suspicion for SFN. Almost half of patients with GWI did not have SFN, so our study does not support SFN as the underlying explanation for all GWI. Although our data cannot provide clinical guidance as to when skin biopsy may be indicated in GWI, work done in fibromyalgia found symptoms of dysautonomia and paresthesias are more specific for SFN and may be useful to help guide medical decision making.³⁹

Veterans with GWI in our clinical sample reported a high burden of clinical symptoms conceivably attributable to ANS dysfunction. This symptom reporting is consistent with that seen in other GWI studies, as well as in other studies of SFN.^{4,5,7-9,14,15,34,38,40} Our clinical sample of veterans with GWI found no differences in the ANS symptom reporting between those with and without SFN. Therefore, our study cannot support SFN alone as accounting for ANS symptom burden in patients with GWI.

Two-thirds of biopsy-confirmed SFN in our clinical sample of veterans with GWI could potentially be explained by established medical conditions. As in other studies of SFN, prediabetes and diabetes represented a plurality (46%). Even after considering hyperlipidemia as a potential explanation, about 21% of SFN cases in veterans with GWI still were deemed idiopathic.

Evidence supports certain environmental agents as causal factors for GWI. Neurotoxicants reportedly related to GWI include pesticides (particularly organophosphates and carbamates), pyridostigmine bromide (used during the Gulf War as a prophylactic agent against the use of chemical weapons), and low levels of the nerve agent sarin from environmental contamination due to chemical weapons detonations.¹ Some of these agents have been implicated in neuropathy as well.^1,28-30 It is biologically plausible that deployment-related exposures could trigger SFN, though the traditional consensus has been that remote exposure to neurotoxic substances is unlikely to produce neuropathy that presents many years after the exposure.⁴¹ In the WRIISC clinical experience, however, veterans often report that their neuropathic symptoms predate the diagnosis of the associated medical conditions, sometimes by decades. It is conceivable that remote exposures may trigger the condition that is then potentiated by ongoing exposures, metabolic factors, and/or other medical conditions. These may perpetuate neuropathic symptoms and the illness experience of affected veterans. Our clinical observation study cannot clarify the extent to which this may be the case. Despite these findings and arguments, an environmental contribution to SFN cannot be discounted, and further research is needed to explore a potential relationship.

Limitations

This study’s conclusions are limited by its observational/retrospective design in a relatively small clinical sample of veterans evaluated at a tertiary referral center for postdeployment exposure-related health concerns. The WRIISC clinical sample is not representative of all GWVs or even of all veterans with GWI, as there is inherent selection bias as to who gets referred to and evaluated at the WRIISC. As with studies based on retrospective chart review, data are reliant on clinical documentation andaccuracy/consistency of the reviewer. Evaluation for SFN with skin biopsy is an invasive procedure and was performed when a high index of clinical suspicion for this condition existed, possibly representing confirmation bias. Therefore, the relatively high prevalence ofbiopsy-confirmed SFN seen in our clinical sample cannot be generalized to GWVs as a whole or even to veterans with GWI.

Assessment of autonomic dysfunction was based on COMPASS 31 symptom reporting by an small subset of the clinical cohort. Symptom reporting may not be reflective of true abnormality in ANS function. Physiologic tests of the ANS were not performed; such studies could more objectively establish whether ANS dysfunction is more prevalent in GWI veterans with SFN.

Evaluation for all potential etiologic/contributory conditions to SFN was not exhaustive. For example, sodium channel gene mutations have been documented to account for up to one-third of all cases of idiopathic SFN.⁴² For those cases in which no compelling etiology was identified, it is plausible that medical explanations for SFN may be found on further investigation.

Clinical assessments at the WRIISC were performed on GWVs ≥ 26 years after their deployment-related exposures. Other conditions/exposures may have occurred in the interim. What is not clear is whether the SFN predated the onset of any of these medical conditions or other putative contributors. This observational study is not able to tease out a temporal association to make a cause-and-effect assessment.

Conclusions

Retrospective analysis of clinical data of veterans evaluated at a specialized center for postdeployment health demonstrated that skin biopsy–confirmed SFN was prevalent, but not ubiquitous, in veterans with GWI. Symptom that may be attributed to ANS dysfunction in this clinical sample was consistent with literature on SFN and with GWI, but we could not definitively attribute ANS symptoms to SFN. Our study does not support the hypothesis that GWI symptoms are solely due to SFN, though it may still be relevant in a subset of veterans with GWI with strongly suggestive clinical features. We were able to identify a potential etiology for SFN in most veterans with GWI. Further investigations are recommended to explore any potential relationship between Gulf War exposures and SFN.