Study Overview

Objective. To determine if there is an association between self-reported English language ability and glycemic control in Latinos with type 2 diabetes.

Design. Descriptive correlational study using data from a larger cross-sectional study.

Setting and participants. 167 adults with diabetes who self-identified as Latino or Hispanic recruited at clinics in the Chicago area from May 2004 to May 2006. The dataset was collected using face-to-face interviews with diabetic patients aged ≥ 18 years. All participants attended clinics affiliated with an academic medical center or physician offices affiliated with a suburban hospital. Patients with type 1 diabetes and those with < 17 points on the Mini-Mental State Examination were excluded. English speaking ability was categorized as speaking English “not at all,” “not well,” “well,” or “very well” based on patient self-report. A multivariable logistic regression model was used to examine the predictive relationship between English language skills and HbA1c levels, with covariates selected if they were significantly correlated with English language ability. The final regression model accounted for age, sex, education, annual income, health insurance status, duration of diabetes, birth in the United States, and years in the United States.

Main outcome measure. HbA1c ≥ 7.0% as captured by chart review.

Main results. Of the 167 patients, 38% reported speaking English very well, 21% reported speaking well, 26% reported speaking not very well, and 14% did not speak English at all. Reflecting immigration-sensitive patterns, patients who spoke English very well were younger and more likely to have graduated high school and have an annual income over $25,000 per year. Comorbidities and complications did not differ by English speaking ability except for diabetic eye disease, which was was more prevalent among those who did not speak English at all (42%, p = 0.04). Whether speaking ability was treated as a continuous or dichotomous variable, HbA1c levels formed a U-shaped curve: those who spoke English very well (odds ratio [OR] 2.32, 95% CI, 1.00–5.41) or not at all (OR 4.11, 95% CI 1.35–12.54) had higher odds of having an elevated HbA1c than those who spoke English well, although this was only statistically significant for those who spoke no English. In adjusted analyses, the U-shaped curve persisted with the highest odds among those who spoke English very well (OR 3.20, 95% CI 1.05–9.79) or not at all (OR 4.95, 95% CI 1.29–18.92).

Conclusion. The relationship between English speaking ability and diabetes management is more complex than previously described. Interventions aimed at improving diabetes outcomes may need to be tailored to specific subgroups within the Latino population.

Commentary

Immigrant health is complex and language is an understudied factor in health transitions of those who migrate for new lives or temporary work. For Latinos, migration abroad was once thought to improve health, but a recent systematic review by Teruya et al [1] suggests that the migration experience has a wide variety of effects on health, many of which can be negative.

The notion that English fluency confers health care benefits is questionable, as the authors state. Those unfamiliar with the acculturation literature might think that English speaking ability is a good marker of acculturation, but recent research on the subject suggests otherwise. Acculturation is a complex phenomenon that cannot be measured or gauged by a single variable [2–5]. Among the many factors influencing acculturation, the migration experience and country of origin will play a major role in acculturation and how it occurs in the arrival country. Health care providers seeking to understand the complexity of acculturation better to improve care for their immigrant patients would benefit from examining the extensive social science literature on the subject. The results of this study suggest that providers should not take for granted someone’s English speaking ability as a marker of acculturation and thus assume that their health outcomes would be equivalent to native born populations.

This study has number of weaknesses. The main concern is that the study did not consider a number of important health service delivery factors. The researchers did not assess for the number of visits the patient had with appropriate interpretation services, whether or not there were language concordant visits between patients and providers (limited English proficiency patients are more likely to form consistent service relationships with language concordant providers [6–10]), or whether the patient had diabetes education classes or individual counseling sessions to facilitate self-management. These service-based factors could potentially explain some of the results seen. The small sample size, age of the data in the study, and failure to distinguish the country of origin of the Latino patients are other weaknesses.

Applications for Clinical Practice

Providers can improve their clinical practice with limited English proficiency Latino patients with diabetes by being more sensitive to the potential effects of language on diabetes outcomes in this population. The results suggest that providers should not assume that a Latino patient’s English language skills mean that they are better at self-managing their diabetes and will have better outcomes. Asking patients about their country of origin and migration experiences may help differentiate the effects of language in concert with other potentially confounding variables that can help elucidate the effects of language on diabetes related outcomes.

            —Allison Squires, PhD, RN

Study Overview

Objective. To determine if there is an association between self-reported English language ability and glycemic control in Latinos with type 2 diabetes.

Design. Descriptive correlational study using data from a larger cross-sectional study.

Setting and participants. 167 adults with diabetes who self-identified as Latino or Hispanic recruited at clinics in the Chicago area from May 2004 to May 2006. The dataset was collected using face-to-face interviews with diabetic patients aged ≥ 18 years. All participants attended clinics affiliated with an academic medical center or physician offices affiliated with a suburban hospital. Patients with type 1 diabetes and those with < 17 points on the Mini-Mental State Examination were excluded. English speaking ability was categorized as speaking English “not at all,” “not well,” “well,” or “very well” based on patient self-report. A multivariable logistic regression model was used to examine the predictive relationship between English language skills and HbA1c levels, with covariates selected if they were significantly correlated with English language ability. The final regression model accounted for age, sex, education, annual income, health insurance status, duration of diabetes, birth in the United States, and years in the United States.

Main outcome measure. HbA1c ≥ 7.0% as captured by chart review.

Main results. Of the 167 patients, 38% reported speaking English very well, 21% reported speaking well, 26% reported speaking not very well, and 14% did not speak English at all. Reflecting immigration-sensitive patterns, patients who spoke English very well were younger and more likely to have graduated high school and have an annual income over $25,000 per year. Comorbidities and complications did not differ by English speaking ability except for diabetic eye disease, which was was more prevalent among those who did not speak English at all (42%, p = 0.04). Whether speaking ability was treated as a continuous or dichotomous variable, HbA1c levels formed a U-shaped curve: those who spoke English very well (odds ratio [OR] 2.32, 95% CI, 1.00–5.41) or not at all (OR 4.11, 95% CI 1.35–12.54) had higher odds of having an elevated HbA1c than those who spoke English well, although this was only statistically significant for those who spoke no English. In adjusted analyses, the U-shaped curve persisted with the highest odds among those who spoke English very well (OR 3.20, 95% CI 1.05–9.79) or not at all (OR 4.95, 95% CI 1.29–18.92).

Conclusion. The relationship between English speaking ability and diabetes management is more complex than previously described. Interventions aimed at improving diabetes outcomes may need to be tailored to specific subgroups within the Latino population.

Commentary

Immigrant health is complex and language is an understudied factor in health transitions of those who migrate for new lives or temporary work. For Latinos, migration abroad was once thought to improve health, but a recent systematic review by Teruya et al [1] suggests that the migration experience has a wide variety of effects on health, many of which can be negative.

The notion that English fluency confers health care benefits is questionable, as the authors state. Those unfamiliar with the acculturation literature might think that English speaking ability is a good marker of acculturation, but recent research on the subject suggests otherwise. Acculturation is a complex phenomenon that cannot be measured or gauged by a single variable [2–5]. Among the many factors influencing acculturation, the migration experience and country of origin will play a major role in acculturation and how it occurs in the arrival country. Health care providers seeking to understand the complexity of acculturation better to improve care for their immigrant patients would benefit from examining the extensive social science literature on the subject. The results of this study suggest that providers should not take for granted someone’s English speaking ability as a marker of acculturation and thus assume that their health outcomes would be equivalent to native born populations.

This study has number of weaknesses. The main concern is that the study did not consider a number of important health service delivery factors. The researchers did not assess for the number of visits the patient had with appropriate interpretation services, whether or not there were language concordant visits between patients and providers (limited English proficiency patients are more likely to form consistent service relationships with language concordant providers [6–10]), or whether the patient had diabetes education classes or individual counseling sessions to facilitate self-management. These service-based factors could potentially explain some of the results seen. The small sample size, age of the data in the study, and failure to distinguish the country of origin of the Latino patients are other weaknesses.

Applications for Clinical Practice

Providers can improve their clinical practice with limited English proficiency Latino patients with diabetes by being more sensitive to the potential effects of language on diabetes outcomes in this population. The results suggest that providers should not assume that a Latino patient’s English language skills mean that they are better at self-managing their diabetes and will have better outcomes. Asking patients about their country of origin and migration experiences may help differentiate the effects of language in concert with other potentially confounding variables that can help elucidate the effects of language on diabetes related outcomes.

            —Allison Squires, PhD, RN

Study Overview

Objective. To determine if there is an association between self-reported English language ability and glycemic control in Latinos with type 2 diabetes.

Design. Descriptive correlational study using data from a larger cross-sectional study.

Setting and participants. 167 adults with diabetes who self-identified as Latino or Hispanic recruited at clinics in the Chicago area from May 2004 to May 2006. The dataset was collected using face-to-face interviews with diabetic patients aged ≥ 18 years. All participants attended clinics affiliated with an academic medical center or physician offices affiliated with a suburban hospital. Patients with type 1 diabetes and those with < 17 points on the Mini-Mental State Examination were excluded. English speaking ability was categorized as speaking English “not at all,” “not well,” “well,” or “very well” based on patient self-report. A multivariable logistic regression model was used to examine the predictive relationship between English language skills and HbA1c levels, with covariates selected if they were significantly correlated with English language ability. The final regression model accounted for age, sex, education, annual income, health insurance status, duration of diabetes, birth in the United States, and years in the United States.

Main outcome measure. HbA1c ≥ 7.0% as captured by chart review.

Main results. Of the 167 patients, 38% reported speaking English very well, 21% reported speaking well, 26% reported speaking not very well, and 14% did not speak English at all. Reflecting immigration-sensitive patterns, patients who spoke English very well were younger and more likely to have graduated high school and have an annual income over $25,000 per year. Comorbidities and complications did not differ by English speaking ability except for diabetic eye disease, which was was more prevalent among those who did not speak English at all (42%, p = 0.04). Whether speaking ability was treated as a continuous or dichotomous variable, HbA1c levels formed a U-shaped curve: those who spoke English very well (odds ratio [OR] 2.32, 95% CI, 1.00–5.41) or not at all (OR 4.11, 95% CI 1.35–12.54) had higher odds of having an elevated HbA1c than those who spoke English well, although this was only statistically significant for those who spoke no English. In adjusted analyses, the U-shaped curve persisted with the highest odds among those who spoke English very well (OR 3.20, 95% CI 1.05–9.79) or not at all (OR 4.95, 95% CI 1.29–18.92).

Conclusion. The relationship between English speaking ability and diabetes management is more complex than previously described. Interventions aimed at improving diabetes outcomes may need to be tailored to specific subgroups within the Latino population.

Commentary

Immigrant health is complex and language is an understudied factor in health transitions of those who migrate for new lives or temporary work. For Latinos, migration abroad was once thought to improve health, but a recent systematic review by Teruya et al [1] suggests that the migration experience has a wide variety of effects on health, many of which can be negative.

The notion that English fluency confers health care benefits is questionable, as the authors state. Those unfamiliar with the acculturation literature might think that English speaking ability is a good marker of acculturation, but recent research on the subject suggests otherwise. Acculturation is a complex phenomenon that cannot be measured or gauged by a single variable [2–5]. Among the many factors influencing acculturation, the migration experience and country of origin will play a major role in acculturation and how it occurs in the arrival country. Health care providers seeking to understand the complexity of acculturation better to improve care for their immigrant patients would benefit from examining the extensive social science literature on the subject. The results of this study suggest that providers should not take for granted someone’s English speaking ability as a marker of acculturation and thus assume that their health outcomes would be equivalent to native born populations.

This study has number of weaknesses. The main concern is that the study did not consider a number of important health service delivery factors. The researchers did not assess for the number of visits the patient had with appropriate interpretation services, whether or not there were language concordant visits between patients and providers (limited English proficiency patients are more likely to form consistent service relationships with language concordant providers [6–10]), or whether the patient had diabetes education classes or individual counseling sessions to facilitate self-management. These service-based factors could potentially explain some of the results seen. The small sample size, age of the data in the study, and failure to distinguish the country of origin of the Latino patients are other weaknesses.

Applications for Clinical Practice

Providers can improve their clinical practice with limited English proficiency Latino patients with diabetes by being more sensitive to the potential effects of language on diabetes outcomes in this population. The results suggest that providers should not assume that a Latino patient’s English language skills mean that they are better at self-managing their diabetes and will have better outcomes. Asking patients about their country of origin and migration experiences may help differentiate the effects of language in concert with other potentially confounding variables that can help elucidate the effects of language on diabetes related outcomes.

            —Allison Squires, PhD, RN

Illustrative case

A 56-year-old man comes to your clinic 3 days after receiving a diagnosis of lower extremity deep vein thrombosis (DVT). He was prescribed warfarin, 5 mg/d, with enoxaparin bridging, 120 mg/d. He has read about post-thrombotic syndrome (PTS) online and is very concerned about this possible side effect. He is asking about using elastic compression stockings (ECS). What should you tell him?

PTS can be a frustrating, debilitating condition. Its clinical features range from minor limb swelling to severe edema and pain, irreversible skin changes, and leg ulcerations.² It occurs in 25% to 50% of patients after DVT.³ Because current PTS treatments are not very effective, prevention is essential.^4,5

Patients are frequently encouraged to wear ECS after DVT to reduce the incidence of PTS by reducing venous hypertension and reflux. These stockings are expensive and uncomfortable. Prior studies suggested that using ECS can cut the incidence of PTS in half.^6,7 However, these were small, single-center studies, and they were not placebo-controlled.^6,7

STUDY SUMMARY: RCT sets aside a common practice

Kahn et al¹ conducted a randomized, placebo-controlled trial of active vs placebo ECS in patients from 24 centers in the United States and Canada who’d had an ultrasound-confirmed proximal DVT (in the popliteal or more proximal deep leg vein) within the previous 14 days. Most patients received standard anticoagulation therapy to treat their DVT (5-10 days of heparin and 3-6 months of warfarin). Patients were excluded if they had received thrombolytics, had arterial claudication, had a life expectancy of <6 months, were unable to put on ECS due to physical disabilities or allergy, or were unable to participate in follow-up visits.

Patients were randomly assigned to wear active (30-40 mm Hg graduated) ECS or identical-looking placebo ECS with <5 mm Hg compression at the ankle for 2 years. Providers, study personnel and statisticians, and patients were all blinded to treatment allocation. Patients were asked to wear the stocking on the affected leg each day from when they woke until they went to bed.

Participants were followed at one, 6, 12, 18, and 24 months. The primary outcome was the cumulative incidence of PTS diagnosed at 6 months or later using Ginsberg’s criteria of ipsilateral pain and swelling of at least 1 month’s duration.⁸ Secondary outcomes included severity of PTS, presence of leg ulcers, recurrence of venous thromboembolism (VTE), death, adverse events, venous valvular reflux, and quality of life (QOL). Outcomes were measured objectively using a validated scale (the Villalta scale) for PTS severity and the 36-item Short Form Health Survey (SF-36) and the Venous Insufficiency Epidemiological and Economic Study Quality of Life (VEINES-QOL) questionnaire to measure QOL.^9-11

There were 409 patients in the ECS group and 394 in the placebo group. Baseline characteristics, including body mass index (BMI), VTE risk factors, and anticoagulation treatment regimens, were similar between groups. The average age was 55.4 years in the study group (standard deviation [SD] ± 15.3 years) and 54.8 years (SD ± 15.8 years) in the place- bo group. Men comprised 62.4% of the active group and 57.9% of the placebo group. Approximately 90% of the participants in both groups were white.

At one month, approximately 95% of participants in both the active and placebo groups used the stockings; at 24 months, a little less than 70% of the participants in both groups continued to use the stockings. The percentage of people who used the stockings for at least 3 days a week was similar across both groups.

The cumulative incidence of PTS during follow-up was 14.2% in the active group vs 12.7% in the placebo group, with a hazard ratio of 1.13 (95% confidence interval [CI], .73-1.76; P=.58). There were no differences in any of the secondary outcomes. Prespecified subgroup analyses found that age, BMI, and severity of DVT had no effect on the outcomes. There was a marginal benefit for ECS for women (P=.047) over men, but this does not likely reflect a true difference because the CIs surrounding the hazard ratios for men and women overlapped and crossed the null value.

WHAT'S NEW: New evidence contradicts previous studies

This placebo-controlled randomized trial found no benefit in using compression stockings to prevent post-thrombotic syndrome.

Two prior studies showed that using 30 to 40 mm Hg ECS decreased the incidence of PTS after proximal DVT.^6,7 However, these were smaller, open-label, single-center studies. This study by Kahn et al¹ was the first placebo-controlled, randomized, multicenter study that used validated instruments to measure PTS and QOL. It found no benefit in using ECS, thus contradicting the results of the prior studies.

There are currently no guidelines or consensus statements for or against the use of ECS after DVT.

CAVEATS: High nonadherence rates might have affected the results

In both groups, adherence to the assigned intervention diminished throughout the study. Overall, approximately 95% of patients reported wearing their stockings at one month; this dropped to just under 70% by 2 years. Theoretically, this could have affected efficacy outcomes. However, the decrease was similar in both groups and represents what is observed in clinical practice. A prespecified per protocol analysis of patients who wore their ECS more regularly found no benefit.

It is possible that a “placebo effect” could explain the lack of difference between groups. However, the placebo stockings provided virtually no compression, and the 2-year cumulative incidence of PTS in both the treatment and placebo groups was similar to that seen in control groups in prior studies.^6,7

Finally, the incidence of PTS in this study was much lower than the 25% to 50% incidence reported in previous studies. Kahn et al¹ suggested that this was because they used more stringent and standardized criteria for PTS than was used in previous research.

CHALLENGES TO IMPLEMENTATION: There are no barriers to ending this practice

We see no challenges to implementation of this recommendation.

Acknowledgement
The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Illustrative case

A 56-year-old man comes to your clinic 3 days after receiving a diagnosis of lower extremity deep vein thrombosis (DVT). He was prescribed warfarin, 5 mg/d, with enoxaparin bridging, 120 mg/d. He has read about post-thrombotic syndrome (PTS) online and is very concerned about this possible side effect. He is asking about using elastic compression stockings (ECS). What should you tell him?

PTS can be a frustrating, debilitating condition. Its clinical features range from minor limb swelling to severe edema and pain, irreversible skin changes, and leg ulcerations.² It occurs in 25% to 50% of patients after DVT.³ Because current PTS treatments are not very effective, prevention is essential.^4,5

Patients are frequently encouraged to wear ECS after DVT to reduce the incidence of PTS by reducing venous hypertension and reflux. These stockings are expensive and uncomfortable. Prior studies suggested that using ECS can cut the incidence of PTS in half.^6,7 However, these were small, single-center studies, and they were not placebo-controlled.^6,7

STUDY SUMMARY: RCT sets aside a common practice

Kahn et al¹ conducted a randomized, placebo-controlled trial of active vs placebo ECS in patients from 24 centers in the United States and Canada who’d had an ultrasound-confirmed proximal DVT (in the popliteal or more proximal deep leg vein) within the previous 14 days. Most patients received standard anticoagulation therapy to treat their DVT (5-10 days of heparin and 3-6 months of warfarin). Patients were excluded if they had received thrombolytics, had arterial claudication, had a life expectancy of <6 months, were unable to put on ECS due to physical disabilities or allergy, or were unable to participate in follow-up visits.

Patients were randomly assigned to wear active (30-40 mm Hg graduated) ECS or identical-looking placebo ECS with <5 mm Hg compression at the ankle for 2 years. Providers, study personnel and statisticians, and patients were all blinded to treatment allocation. Patients were asked to wear the stocking on the affected leg each day from when they woke until they went to bed.

Participants were followed at one, 6, 12, 18, and 24 months. The primary outcome was the cumulative incidence of PTS diagnosed at 6 months or later using Ginsberg’s criteria of ipsilateral pain and swelling of at least 1 month’s duration.⁸ Secondary outcomes included severity of PTS, presence of leg ulcers, recurrence of venous thromboembolism (VTE), death, adverse events, venous valvular reflux, and quality of life (QOL). Outcomes were measured objectively using a validated scale (the Villalta scale) for PTS severity and the 36-item Short Form Health Survey (SF-36) and the Venous Insufficiency Epidemiological and Economic Study Quality of Life (VEINES-QOL) questionnaire to measure QOL.^9-11

There were 409 patients in the ECS group and 394 in the placebo group. Baseline characteristics, including body mass index (BMI), VTE risk factors, and anticoagulation treatment regimens, were similar between groups. The average age was 55.4 years in the study group (standard deviation [SD] ± 15.3 years) and 54.8 years (SD ± 15.8 years) in the place- bo group. Men comprised 62.4% of the active group and 57.9% of the placebo group. Approximately 90% of the participants in both groups were white.

At one month, approximately 95% of participants in both the active and placebo groups used the stockings; at 24 months, a little less than 70% of the participants in both groups continued to use the stockings. The percentage of people who used the stockings for at least 3 days a week was similar across both groups.

The cumulative incidence of PTS during follow-up was 14.2% in the active group vs 12.7% in the placebo group, with a hazard ratio of 1.13 (95% confidence interval [CI], .73-1.76; P=.58). There were no differences in any of the secondary outcomes. Prespecified subgroup analyses found that age, BMI, and severity of DVT had no effect on the outcomes. There was a marginal benefit for ECS for women (P=.047) over men, but this does not likely reflect a true difference because the CIs surrounding the hazard ratios for men and women overlapped and crossed the null value.

WHAT'S NEW: New evidence contradicts previous studies

This placebo-controlled randomized trial found no benefit in using compression stockings to prevent post-thrombotic syndrome.

Two prior studies showed that using 30 to 40 mm Hg ECS decreased the incidence of PTS after proximal DVT.^6,7 However, these were smaller, open-label, single-center studies. This study by Kahn et al¹ was the first placebo-controlled, randomized, multicenter study that used validated instruments to measure PTS and QOL. It found no benefit in using ECS, thus contradicting the results of the prior studies.

There are currently no guidelines or consensus statements for or against the use of ECS after DVT.

CAVEATS: High nonadherence rates might have affected the results

In both groups, adherence to the assigned intervention diminished throughout the study. Overall, approximately 95% of patients reported wearing their stockings at one month; this dropped to just under 70% by 2 years. Theoretically, this could have affected efficacy outcomes. However, the decrease was similar in both groups and represents what is observed in clinical practice. A prespecified per protocol analysis of patients who wore their ECS more regularly found no benefit.

It is possible that a “placebo effect” could explain the lack of difference between groups. However, the placebo stockings provided virtually no compression, and the 2-year cumulative incidence of PTS in both the treatment and placebo groups was similar to that seen in control groups in prior studies.^6,7

Finally, the incidence of PTS in this study was much lower than the 25% to 50% incidence reported in previous studies. Kahn et al¹ suggested that this was because they used more stringent and standardized criteria for PTS than was used in previous research.

CHALLENGES TO IMPLEMENTATION: There are no barriers to ending this practice

We see no challenges to implementation of this recommendation.

Acknowledgement
The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

Illustrative case

A 56-year-old man comes to your clinic 3 days after receiving a diagnosis of lower extremity deep vein thrombosis (DVT). He was prescribed warfarin, 5 mg/d, with enoxaparin bridging, 120 mg/d. He has read about post-thrombotic syndrome (PTS) online and is very concerned about this possible side effect. He is asking about using elastic compression stockings (ECS). What should you tell him?

PTS can be a frustrating, debilitating condition. Its clinical features range from minor limb swelling to severe edema and pain, irreversible skin changes, and leg ulcerations.² It occurs in 25% to 50% of patients after DVT.³ Because current PTS treatments are not very effective, prevention is essential.^4,5

Patients are frequently encouraged to wear ECS after DVT to reduce the incidence of PTS by reducing venous hypertension and reflux. These stockings are expensive and uncomfortable. Prior studies suggested that using ECS can cut the incidence of PTS in half.^6,7 However, these were small, single-center studies, and they were not placebo-controlled.^6,7

STUDY SUMMARY: RCT sets aside a common practice

Kahn et al¹ conducted a randomized, placebo-controlled trial of active vs placebo ECS in patients from 24 centers in the United States and Canada who’d had an ultrasound-confirmed proximal DVT (in the popliteal or more proximal deep leg vein) within the previous 14 days. Most patients received standard anticoagulation therapy to treat their DVT (5-10 days of heparin and 3-6 months of warfarin). Patients were excluded if they had received thrombolytics, had arterial claudication, had a life expectancy of <6 months, were unable to put on ECS due to physical disabilities or allergy, or were unable to participate in follow-up visits.

Patients were randomly assigned to wear active (30-40 mm Hg graduated) ECS or identical-looking placebo ECS with <5 mm Hg compression at the ankle for 2 years. Providers, study personnel and statisticians, and patients were all blinded to treatment allocation. Patients were asked to wear the stocking on the affected leg each day from when they woke until they went to bed.

Participants were followed at one, 6, 12, 18, and 24 months. The primary outcome was the cumulative incidence of PTS diagnosed at 6 months or later using Ginsberg’s criteria of ipsilateral pain and swelling of at least 1 month’s duration.⁸ Secondary outcomes included severity of PTS, presence of leg ulcers, recurrence of venous thromboembolism (VTE), death, adverse events, venous valvular reflux, and quality of life (QOL). Outcomes were measured objectively using a validated scale (the Villalta scale) for PTS severity and the 36-item Short Form Health Survey (SF-36) and the Venous Insufficiency Epidemiological and Economic Study Quality of Life (VEINES-QOL) questionnaire to measure QOL.^9-11

There were 409 patients in the ECS group and 394 in the placebo group. Baseline characteristics, including body mass index (BMI), VTE risk factors, and anticoagulation treatment regimens, were similar between groups. The average age was 55.4 years in the study group (standard deviation [SD] ± 15.3 years) and 54.8 years (SD ± 15.8 years) in the place- bo group. Men comprised 62.4% of the active group and 57.9% of the placebo group. Approximately 90% of the participants in both groups were white.

At one month, approximately 95% of participants in both the active and placebo groups used the stockings; at 24 months, a little less than 70% of the participants in both groups continued to use the stockings. The percentage of people who used the stockings for at least 3 days a week was similar across both groups.

The cumulative incidence of PTS during follow-up was 14.2% in the active group vs 12.7% in the placebo group, with a hazard ratio of 1.13 (95% confidence interval [CI], .73-1.76; P=.58). There were no differences in any of the secondary outcomes. Prespecified subgroup analyses found that age, BMI, and severity of DVT had no effect on the outcomes. There was a marginal benefit for ECS for women (P=.047) over men, but this does not likely reflect a true difference because the CIs surrounding the hazard ratios for men and women overlapped and crossed the null value.

WHAT'S NEW: New evidence contradicts previous studies

This placebo-controlled randomized trial found no benefit in using compression stockings to prevent post-thrombotic syndrome.

Two prior studies showed that using 30 to 40 mm Hg ECS decreased the incidence of PTS after proximal DVT.^6,7 However, these were smaller, open-label, single-center studies. This study by Kahn et al¹ was the first placebo-controlled, randomized, multicenter study that used validated instruments to measure PTS and QOL. It found no benefit in using ECS, thus contradicting the results of the prior studies.

There are currently no guidelines or consensus statements for or against the use of ECS after DVT.

CAVEATS: High nonadherence rates might have affected the results

In both groups, adherence to the assigned intervention diminished throughout the study. Overall, approximately 95% of patients reported wearing their stockings at one month; this dropped to just under 70% by 2 years. Theoretically, this could have affected efficacy outcomes. However, the decrease was similar in both groups and represents what is observed in clinical practice. A prespecified per protocol analysis of patients who wore their ECS more regularly found no benefit.

It is possible that a “placebo effect” could explain the lack of difference between groups. However, the placebo stockings provided virtually no compression, and the 2-year cumulative incidence of PTS in both the treatment and placebo groups was similar to that seen in control groups in prior studies.^6,7

Finally, the incidence of PTS in this study was much lower than the 25% to 50% incidence reported in previous studies. Kahn et al¹ suggested that this was because they used more stringent and standardized criteria for PTS than was used in previous research.

CHALLENGES TO IMPLEMENTATION: There are no barriers to ending this practice

We see no challenges to implementation of this recommendation.

Acknowledgement
The PURLs Surveillance System was supported in part by Grant Number UL1RR024999 from the National Center For Research Resources, a Clinical Translational Science Award to the University of Chicago. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.

To the Editor: We read with interest the article by Drs. Buitrago et al in the May 2014 issue of Cleveland Clinic Journal of Medicine, “Syncope during a pharmacologic nuclear stress test.”¹ It highlights a known, serious interaction between adenosine and dipyridamole (the latter contained in the aspirin-dipyridamole combination Aggrenox) and associated asystole in patients undergoing pharmacologic cardiac stress testing. This interaction is known in the cardiology literature, as it was noted in the current guidelines for pharmacologic stress testing.² However, I would like to discuss a few points with the authors for a better understanding of the case.

First, the underlying rhythm before the development of complete atrioventricular (AV) dissociation and asystole was significant for second-degree AV block (Mobitz type I, Wenckebach). Second- or third-degree AV block is considered a contraindication to adenosine because of the risk of exacerbating these conditions. This underlying AV nodal disease made dipyridamole not the only culprit. In addition, the patient had been on two agents (labetalol and clonidine) that have AV nodal-blocking properties. Electrolyte imbalances such as hypokalemia, hypomagnesemia, and hypocalcemia are another reason for delayed conduction and PR prolongation, and electrolyte levels should be checked and corrected properly before the stress test or coronary angiography. It would have been helpful if the authors had discussed these points for a better understanding of the drug-drug interaction.

Because of the increasing trend to admit patients with chest pain to observation units to rule out myocardial infarction, the case has a valuable teaching point, especially for hospitalists and emergency physicians in charge of patients admitted with chest pain.³ Since cardiologists rarely get involved in the care of these patients, careful review of medications before scheduling stress testing is of ultimate importance and should be emphasized in the discussion.

Lastly, the number of combined medications that are available commercially is increasing, which puts patients at higher risk of drug interactions. Hospitalists and internists taking care of patients, especially elderly patients, admitted from nursing homes and taking multiple medications should pay extra attention when reviewing medications with brand names.^4,5 Furthermore, a 12-lead electrocardiogram should be reviewed, with special attention to the PR interval and QT segment. A pharmacy consultation could be valuable, especially in patients taking multiple drugs.⁶

To the Editor: We read with interest the article by Drs. Buitrago et al in the May 2014 issue of Cleveland Clinic Journal of Medicine, “Syncope during a pharmacologic nuclear stress test.”¹ It highlights a known, serious interaction between adenosine and dipyridamole (the latter contained in the aspirin-dipyridamole combination Aggrenox) and associated asystole in patients undergoing pharmacologic cardiac stress testing. This interaction is known in the cardiology literature, as it was noted in the current guidelines for pharmacologic stress testing.² However, I would like to discuss a few points with the authors for a better understanding of the case.

First, the underlying rhythm before the development of complete atrioventricular (AV) dissociation and asystole was significant for second-degree AV block (Mobitz type I, Wenckebach). Second- or third-degree AV block is considered a contraindication to adenosine because of the risk of exacerbating these conditions. This underlying AV nodal disease made dipyridamole not the only culprit. In addition, the patient had been on two agents (labetalol and clonidine) that have AV nodal-blocking properties. Electrolyte imbalances such as hypokalemia, hypomagnesemia, and hypocalcemia are another reason for delayed conduction and PR prolongation, and electrolyte levels should be checked and corrected properly before the stress test or coronary angiography. It would have been helpful if the authors had discussed these points for a better understanding of the drug-drug interaction.

Because of the increasing trend to admit patients with chest pain to observation units to rule out myocardial infarction, the case has a valuable teaching point, especially for hospitalists and emergency physicians in charge of patients admitted with chest pain.³ Since cardiologists rarely get involved in the care of these patients, careful review of medications before scheduling stress testing is of ultimate importance and should be emphasized in the discussion.

Lastly, the number of combined medications that are available commercially is increasing, which puts patients at higher risk of drug interactions. Hospitalists and internists taking care of patients, especially elderly patients, admitted from nursing homes and taking multiple medications should pay extra attention when reviewing medications with brand names.^4,5 Furthermore, a 12-lead electrocardiogram should be reviewed, with special attention to the PR interval and QT segment. A pharmacy consultation could be valuable, especially in patients taking multiple drugs.⁶

To the Editor: We read with interest the article by Drs. Buitrago et al in the May 2014 issue of Cleveland Clinic Journal of Medicine, “Syncope during a pharmacologic nuclear stress test.”¹ It highlights a known, serious interaction between adenosine and dipyridamole (the latter contained in the aspirin-dipyridamole combination Aggrenox) and associated asystole in patients undergoing pharmacologic cardiac stress testing. This interaction is known in the cardiology literature, as it was noted in the current guidelines for pharmacologic stress testing.² However, I would like to discuss a few points with the authors for a better understanding of the case.

First, the underlying rhythm before the development of complete atrioventricular (AV) dissociation and asystole was significant for second-degree AV block (Mobitz type I, Wenckebach). Second- or third-degree AV block is considered a contraindication to adenosine because of the risk of exacerbating these conditions. This underlying AV nodal disease made dipyridamole not the only culprit. In addition, the patient had been on two agents (labetalol and clonidine) that have AV nodal-blocking properties. Electrolyte imbalances such as hypokalemia, hypomagnesemia, and hypocalcemia are another reason for delayed conduction and PR prolongation, and electrolyte levels should be checked and corrected properly before the stress test or coronary angiography. It would have been helpful if the authors had discussed these points for a better understanding of the drug-drug interaction.

Because of the increasing trend to admit patients with chest pain to observation units to rule out myocardial infarction, the case has a valuable teaching point, especially for hospitalists and emergency physicians in charge of patients admitted with chest pain.³ Since cardiologists rarely get involved in the care of these patients, careful review of medications before scheduling stress testing is of ultimate importance and should be emphasized in the discussion.

Lastly, the number of combined medications that are available commercially is increasing, which puts patients at higher risk of drug interactions. Hospitalists and internists taking care of patients, especially elderly patients, admitted from nursing homes and taking multiple medications should pay extra attention when reviewing medications with brand names.^4,5 Furthermore, a 12-lead electrocardiogram should be reviewed, with special attention to the PR interval and QT segment. A pharmacy consultation could be valuable, especially in patients taking multiple drugs.⁶

In Reply: We appreciate the interest and comments of Dr. Alraies. We would like to clarify that the patient’s baseline electrocardiogram before the nuclear stress test was normal. Second-degree atrioventricular (AV) block (Mobitz type I) was evident only during adenosine infusion before ventricular asystole. The patient was on two AV nodal blockers (labetalol and clonidine) but had no underlying conduction disease. There is no contraindication to continuing these agents before pharmacologic stress testing. In addition, the patient’s electrolyte levels were within normal ranges before testing.

We agree that the valuable teaching point for clinicians is to appreciate the contraindication to and consequences of the use of dipyridamole-containing oral medications and either adenosine or regadenoson during pharmacologic stress testing. As Dr. Alraies points out, most cardiologists may be familiar with this interaction, but a large proportion of stress tests are ordered by emergency room physicians, internists, and hospitalists who are not. Still, the overall incidence of side effects with pharmacologic stress testing is very low and comparable to that with exercise testing, with safety enhanced by following the American Society of Nuclear Cardiology (ASNC) guidelines for performing stress myocardial perfusion imaging.¹ Avoidance of this interaction may be enhanced through education, but also by using checklists and building notifications into the electronic medical record when ordering pharmacologic stress testing. Of note, according to the ASNC guidelines, the use of intravenous dipyridamole as a stress agent is a safe alternative for pharmacologic stress testing in patients taking oral dipyridamole-containing medications.

In Reply: We appreciate the interest and comments of Dr. Alraies. We would like to clarify that the patient’s baseline electrocardiogram before the nuclear stress test was normal. Second-degree atrioventricular (AV) block (Mobitz type I) was evident only during adenosine infusion before ventricular asystole. The patient was on two AV nodal blockers (labetalol and clonidine) but had no underlying conduction disease. There is no contraindication to continuing these agents before pharmacologic stress testing. In addition, the patient’s electrolyte levels were within normal ranges before testing.

We agree that the valuable teaching point for clinicians is to appreciate the contraindication to and consequences of the use of dipyridamole-containing oral medications and either adenosine or regadenoson during pharmacologic stress testing. As Dr. Alraies points out, most cardiologists may be familiar with this interaction, but a large proportion of stress tests are ordered by emergency room physicians, internists, and hospitalists who are not. Still, the overall incidence of side effects with pharmacologic stress testing is very low and comparable to that with exercise testing, with safety enhanced by following the American Society of Nuclear Cardiology (ASNC) guidelines for performing stress myocardial perfusion imaging.¹ Avoidance of this interaction may be enhanced through education, but also by using checklists and building notifications into the electronic medical record when ordering pharmacologic stress testing. Of note, according to the ASNC guidelines, the use of intravenous dipyridamole as a stress agent is a safe alternative for pharmacologic stress testing in patients taking oral dipyridamole-containing medications.

In Reply: We appreciate the interest and comments of Dr. Alraies. We would like to clarify that the patient’s baseline electrocardiogram before the nuclear stress test was normal. Second-degree atrioventricular (AV) block (Mobitz type I) was evident only during adenosine infusion before ventricular asystole. The patient was on two AV nodal blockers (labetalol and clonidine) but had no underlying conduction disease. There is no contraindication to continuing these agents before pharmacologic stress testing. In addition, the patient’s electrolyte levels were within normal ranges before testing.

We agree that the valuable teaching point for clinicians is to appreciate the contraindication to and consequences of the use of dipyridamole-containing oral medications and either adenosine or regadenoson during pharmacologic stress testing. As Dr. Alraies points out, most cardiologists may be familiar with this interaction, but a large proportion of stress tests are ordered by emergency room physicians, internists, and hospitalists who are not. Still, the overall incidence of side effects with pharmacologic stress testing is very low and comparable to that with exercise testing, with safety enhanced by following the American Society of Nuclear Cardiology (ASNC) guidelines for performing stress myocardial perfusion imaging.¹ Avoidance of this interaction may be enhanced through education, but also by using checklists and building notifications into the electronic medical record when ordering pharmacologic stress testing. Of note, according to the ASNC guidelines, the use of intravenous dipyridamole as a stress agent is a safe alternative for pharmacologic stress testing in patients taking oral dipyridamole-containing medications.