User login
Endovascular lower-extremity revascularization by vascular specialists resulted in more transfusions and ICU use, as well as a longer hospital stay, more repeat revascularization procedures and amputations, and higher costs, compared with the same procedures performed by interventional radiologists, according to the results of a large database analysis of the Medicare 5% Standard Analytical Files.
Dr. Abdul M. Zafar of the Vascular Disease Research Center, Brown University Alpert Medical School, Providence, and his colleagues extracted claims for endovascularlower extremity revascularization (ELER) index procedures using Current Procedural Terminology (CPT) codes in the years 2005 through 2007.
They recorded the unique IDs of the 15,455 patients who underwent these procedures and for whom an inpatient or outpatient claim corresponding to the index procedure claim was available. This comprised all claims for percutaneous angioplasty, atherectomy, and stent implantation of lower extremity arteries.
The self-designated specialty code of the physician performing the procedure was used to determine specialist type: vascular surgeons (VS), interventional radiologists (IR); interventional cardiologists (IC); and "other."
After excluding patients who might have undergone hybrid procedures and those who underwent thrombolysis procedures in the same year as the index procedure, the final sample of 14,608 patients was analyzed for the following outcomes: length of hospital stay; use of intensive care unit (ICU) services; transfusions; in-hospital mortality; and repeat intervention (defined as any ELER, open lower-extremity revascularization, or amputation of the lower extremity).
The researchers built risk-adjusted logistic regression models by using maximum-likelihood estimates to compare various patient outcomes across different specialties, and they used a linear regression model employing ordinary least squares to analyze length of stay. Cost analysis was performed using a linear regression model based on the least-squares approach. All models were adjusted for age, sex, race, admission type (emergency or ambulatory), and other comorbidities using the Elixhauser comorbidities software available from the Agency for Healthcare Research and Quality. Other regression models that included the International Classification of Disease-9 code for disease severity were developed, but were less predictive and had lower R2 values, according to the authors. All logistic regression models passed the goodness-of-fit test at the 99% confidence level. In addition, linear regression models were also found to be satisfactory at the same level, according to the authors.
Within the study, there were 3,565 index procedures done by IRs; 5,489 by ICs; 5,358 by VSs; and 196 performed by "other" specialties (J. Vasc. Interven. Radiol 2012;23:3-9).
Dr. Zafar and his colleagues reported that VS outcomes were significantly worse than when the procedures were performed by an IR or IC. IRs had a 32% lower likelihood of ICU use (P less than .001) and a 37% lower likelihood of repeat lower-extremity revascularization or amputation (P less than .001) compared with VS. "Although statistical significance was not reached, both transfusion use and in-hospital mortality were 19% less likely after IRs performed procedures compared with VSs (P = .113 and P =.351, respectively)."
"Vascular surgeons were the only specialists with post-index procedure length of stay exceeding 3 days, significantly longer than observed for other specialties. The adjusted average 1-year costs per index procedure were 9% greater for vascular surgeons than for interventional radiologists ($19,012 vs. $17,640)," they said.
"The reasons for worse outcomes among VSs are not known, but may be related to insufficient training in catheter-based interventions as a result of the extensive time learning and practicing open surgical procedure compared with IRs and ICs, whose focus is catheter interventions," they said. "Medicare data indicate that patients who need lower-extremity endovascular revascularization services experience shorter hospital stays, require less transfusions and ICU services, have lower in-hospital mortality rates, and have much less chance of a subsequent revascularization or amputation within one year if treated by an IR rather than a VS," they concluded.
The study was supported by a grant from the Society of Interventional Radiology and internal funds from the Vascular Disease Research Center. One authors received research funding from Cordis/Johnson&Johnson, and Abbott Vascular and consulting for Microvention/Terumo.
I note that the work was supported by a grant from the Society of Interventional Radiologists (SIR), and that the corresponding author is the current SIR President. The manuscript itself is so methodologically flawed ....its manipulated data alone would be enough to call it into question. The introductory paragraph states the authors’ message is that vascular surgeons offer endovascular procedures related to economic greed. This repugnant accusation with respect to monetary concerns motivating the practice of vascular surgeons (VS) represents a nadir of professional conduct.
The theme is continued in the Discussion where the authors make their ridiculous contention that vascular surgeons offering both open and endovascular procedures represents an inherent conflict of interest... against Medicare beneficiaries to capture more revenue. [In fact,] since VS are the only specialists offering both open and endovascular options to patients, it must be apparent that VS make treatment recommendations without prejudice or bias to any particular mode of intervention.
As in other vascular territories, treatment of leg occlusive disease has largely shifted from open to endovascular procedures; and Interventional Radiologists (IR) "market share" for such procedures declined from 67% in 1996 to less than 20% in 2006. As in other vascular territories, patients are best served when VS who provide comprehensive care and longitudinal follow-up, also manage their intervention irrespective of its nature. It is certainly true that SVS and others engineered a retraining of the VS workforce over the past 15 years, and that in terms of endovascular training, the paradigm has shifted. VS trainees have had mandated endovascular training with case numbers requirements for a decade, and their experience in this realm now far exceeds that of IR fellows.
The selection of endpoints, such as transfusion, use of intensive care services, and length of stay are often irrelevant (in hospitalized patients) to an endovascular procedure per se; rather, they reflect the overall complexity and/or complicating patient comorbidities that may dominate the clinical picture. Nor can the Medicare database used in this study discern the temporal relationship of endpoints such as ICU stay to the vascular intervention. In consideration of the mortality endpoint, they repeatedly call attention to the "19% higher mortality with vascular surgeons," but note this difference was not significant. In addition, the selection of a subsequent revascularization or amputation as a surrogate for procedural quality introduces the single glaring flaw in this manuscript, viz. the failure to include specifics of the indication for vascular intervention. These data are available in the Medicare database, but we are told that inclusion of such data in regression models were "less predictive and had a low R-square values." The authors manipulating of the data with what we consider illogical surrogates for procedural expertise may be in response to prior work, using the NIS database that demonstrated both procedural mortality and iatrogenic arterial injury were significantly higher for IR and IC as compared to VS. Indeed this earlier report also utilized appropriate risk adjusted multivariate analyses, which is inherently lacking in the JVIR article. Furthermore, it is well documented that VS are significantly more likely to treat patients with rest pain and tissue loss, as compared to claudicants.
It is also documented and intuitively logical that virtually all of the endpoints that the authors consider, including mortality, LOS, amputation rates, and resource consumption are significantly higher in patients treated for limb-threatening ischemia as opposed to claudication.
The analysis was adjusted only for age, sex, race, admission type, and a general co-morbidity measure (Elixhauser method), which accounts for 30 variables including inconsequential considerations, such as weight loss and hypothyroidism, but not for the indication for the procedure. Thus when used as a total score, this method includes a great deal of ‘noise,’ making it largely irrelevant to treatment outcomes in patients with PAD. Such indices were developed to be used at administrative levels in comparing hospital or health systems rather than individual patient outcomes.
Finally, the use of the endpoint of repeat revascularization or any amputation as the surrogate for quality of the index procedure is patently absurd. A digital or forefoot amputation is often performed subsequent to a lower extremity revascularization in patients with tissue loss. Furthermore, secondary interventions after infrainguinal endovascular procedures are commonly required to maintain secondary (reported ranges 15-30%) patency.
To those with any knowledge of vascular disease, this article will be seen as self-serving and scientifically flawed.
Dr. Richard P. Cambria, President, Society for Vascular Surgery
I note that the work was supported by a grant from the Society of Interventional Radiologists (SIR), and that the corresponding author is the current SIR President. The manuscript itself is so methodologically flawed ....its manipulated data alone would be enough to call it into question. The introductory paragraph states the authors’ message is that vascular surgeons offer endovascular procedures related to economic greed. This repugnant accusation with respect to monetary concerns motivating the practice of vascular surgeons (VS) represents a nadir of professional conduct.
The theme is continued in the Discussion where the authors make their ridiculous contention that vascular surgeons offering both open and endovascular procedures represents an inherent conflict of interest... against Medicare beneficiaries to capture more revenue. [In fact,] since VS are the only specialists offering both open and endovascular options to patients, it must be apparent that VS make treatment recommendations without prejudice or bias to any particular mode of intervention.
As in other vascular territories, treatment of leg occlusive disease has largely shifted from open to endovascular procedures; and Interventional Radiologists (IR) "market share" for such procedures declined from 67% in 1996 to less than 20% in 2006. As in other vascular territories, patients are best served when VS who provide comprehensive care and longitudinal follow-up, also manage their intervention irrespective of its nature. It is certainly true that SVS and others engineered a retraining of the VS workforce over the past 15 years, and that in terms of endovascular training, the paradigm has shifted. VS trainees have had mandated endovascular training with case numbers requirements for a decade, and their experience in this realm now far exceeds that of IR fellows.
The selection of endpoints, such as transfusion, use of intensive care services, and length of stay are often irrelevant (in hospitalized patients) to an endovascular procedure per se; rather, they reflect the overall complexity and/or complicating patient comorbidities that may dominate the clinical picture. Nor can the Medicare database used in this study discern the temporal relationship of endpoints such as ICU stay to the vascular intervention. In consideration of the mortality endpoint, they repeatedly call attention to the "19% higher mortality with vascular surgeons," but note this difference was not significant. In addition, the selection of a subsequent revascularization or amputation as a surrogate for procedural quality introduces the single glaring flaw in this manuscript, viz. the failure to include specifics of the indication for vascular intervention. These data are available in the Medicare database, but we are told that inclusion of such data in regression models were "less predictive and had a low R-square values." The authors manipulating of the data with what we consider illogical surrogates for procedural expertise may be in response to prior work, using the NIS database that demonstrated both procedural mortality and iatrogenic arterial injury were significantly higher for IR and IC as compared to VS. Indeed this earlier report also utilized appropriate risk adjusted multivariate analyses, which is inherently lacking in the JVIR article. Furthermore, it is well documented that VS are significantly more likely to treat patients with rest pain and tissue loss, as compared to claudicants.
It is also documented and intuitively logical that virtually all of the endpoints that the authors consider, including mortality, LOS, amputation rates, and resource consumption are significantly higher in patients treated for limb-threatening ischemia as opposed to claudication.
The analysis was adjusted only for age, sex, race, admission type, and a general co-morbidity measure (Elixhauser method), which accounts for 30 variables including inconsequential considerations, such as weight loss and hypothyroidism, but not for the indication for the procedure. Thus when used as a total score, this method includes a great deal of ‘noise,’ making it largely irrelevant to treatment outcomes in patients with PAD. Such indices were developed to be used at administrative levels in comparing hospital or health systems rather than individual patient outcomes.
Finally, the use of the endpoint of repeat revascularization or any amputation as the surrogate for quality of the index procedure is patently absurd. A digital or forefoot amputation is often performed subsequent to a lower extremity revascularization in patients with tissue loss. Furthermore, secondary interventions after infrainguinal endovascular procedures are commonly required to maintain secondary (reported ranges 15-30%) patency.
To those with any knowledge of vascular disease, this article will be seen as self-serving and scientifically flawed.
Dr. Richard P. Cambria, President, Society for Vascular Surgery
I note that the work was supported by a grant from the Society of Interventional Radiologists (SIR), and that the corresponding author is the current SIR President. The manuscript itself is so methodologically flawed ....its manipulated data alone would be enough to call it into question. The introductory paragraph states the authors’ message is that vascular surgeons offer endovascular procedures related to economic greed. This repugnant accusation with respect to monetary concerns motivating the practice of vascular surgeons (VS) represents a nadir of professional conduct.
The theme is continued in the Discussion where the authors make their ridiculous contention that vascular surgeons offering both open and endovascular procedures represents an inherent conflict of interest... against Medicare beneficiaries to capture more revenue. [In fact,] since VS are the only specialists offering both open and endovascular options to patients, it must be apparent that VS make treatment recommendations without prejudice or bias to any particular mode of intervention.
As in other vascular territories, treatment of leg occlusive disease has largely shifted from open to endovascular procedures; and Interventional Radiologists (IR) "market share" for such procedures declined from 67% in 1996 to less than 20% in 2006. As in other vascular territories, patients are best served when VS who provide comprehensive care and longitudinal follow-up, also manage their intervention irrespective of its nature. It is certainly true that SVS and others engineered a retraining of the VS workforce over the past 15 years, and that in terms of endovascular training, the paradigm has shifted. VS trainees have had mandated endovascular training with case numbers requirements for a decade, and their experience in this realm now far exceeds that of IR fellows.
The selection of endpoints, such as transfusion, use of intensive care services, and length of stay are often irrelevant (in hospitalized patients) to an endovascular procedure per se; rather, they reflect the overall complexity and/or complicating patient comorbidities that may dominate the clinical picture. Nor can the Medicare database used in this study discern the temporal relationship of endpoints such as ICU stay to the vascular intervention. In consideration of the mortality endpoint, they repeatedly call attention to the "19% higher mortality with vascular surgeons," but note this difference was not significant. In addition, the selection of a subsequent revascularization or amputation as a surrogate for procedural quality introduces the single glaring flaw in this manuscript, viz. the failure to include specifics of the indication for vascular intervention. These data are available in the Medicare database, but we are told that inclusion of such data in regression models were "less predictive and had a low R-square values." The authors manipulating of the data with what we consider illogical surrogates for procedural expertise may be in response to prior work, using the NIS database that demonstrated both procedural mortality and iatrogenic arterial injury were significantly higher for IR and IC as compared to VS. Indeed this earlier report also utilized appropriate risk adjusted multivariate analyses, which is inherently lacking in the JVIR article. Furthermore, it is well documented that VS are significantly more likely to treat patients with rest pain and tissue loss, as compared to claudicants.
It is also documented and intuitively logical that virtually all of the endpoints that the authors consider, including mortality, LOS, amputation rates, and resource consumption are significantly higher in patients treated for limb-threatening ischemia as opposed to claudication.
The analysis was adjusted only for age, sex, race, admission type, and a general co-morbidity measure (Elixhauser method), which accounts for 30 variables including inconsequential considerations, such as weight loss and hypothyroidism, but not for the indication for the procedure. Thus when used as a total score, this method includes a great deal of ‘noise,’ making it largely irrelevant to treatment outcomes in patients with PAD. Such indices were developed to be used at administrative levels in comparing hospital or health systems rather than individual patient outcomes.
Finally, the use of the endpoint of repeat revascularization or any amputation as the surrogate for quality of the index procedure is patently absurd. A digital or forefoot amputation is often performed subsequent to a lower extremity revascularization in patients with tissue loss. Furthermore, secondary interventions after infrainguinal endovascular procedures are commonly required to maintain secondary (reported ranges 15-30%) patency.
To those with any knowledge of vascular disease, this article will be seen as self-serving and scientifically flawed.
Dr. Richard P. Cambria, President, Society for Vascular Surgery
Endovascular lower-extremity revascularization by vascular specialists resulted in more transfusions and ICU use, as well as a longer hospital stay, more repeat revascularization procedures and amputations, and higher costs, compared with the same procedures performed by interventional radiologists, according to the results of a large database analysis of the Medicare 5% Standard Analytical Files.
Dr. Abdul M. Zafar of the Vascular Disease Research Center, Brown University Alpert Medical School, Providence, and his colleagues extracted claims for endovascularlower extremity revascularization (ELER) index procedures using Current Procedural Terminology (CPT) codes in the years 2005 through 2007.
They recorded the unique IDs of the 15,455 patients who underwent these procedures and for whom an inpatient or outpatient claim corresponding to the index procedure claim was available. This comprised all claims for percutaneous angioplasty, atherectomy, and stent implantation of lower extremity arteries.
The self-designated specialty code of the physician performing the procedure was used to determine specialist type: vascular surgeons (VS), interventional radiologists (IR); interventional cardiologists (IC); and "other."
After excluding patients who might have undergone hybrid procedures and those who underwent thrombolysis procedures in the same year as the index procedure, the final sample of 14,608 patients was analyzed for the following outcomes: length of hospital stay; use of intensive care unit (ICU) services; transfusions; in-hospital mortality; and repeat intervention (defined as any ELER, open lower-extremity revascularization, or amputation of the lower extremity).
The researchers built risk-adjusted logistic regression models by using maximum-likelihood estimates to compare various patient outcomes across different specialties, and they used a linear regression model employing ordinary least squares to analyze length of stay. Cost analysis was performed using a linear regression model based on the least-squares approach. All models were adjusted for age, sex, race, admission type (emergency or ambulatory), and other comorbidities using the Elixhauser comorbidities software available from the Agency for Healthcare Research and Quality. Other regression models that included the International Classification of Disease-9 code for disease severity were developed, but were less predictive and had lower R2 values, according to the authors. All logistic regression models passed the goodness-of-fit test at the 99% confidence level. In addition, linear regression models were also found to be satisfactory at the same level, according to the authors.
Within the study, there were 3,565 index procedures done by IRs; 5,489 by ICs; 5,358 by VSs; and 196 performed by "other" specialties (J. Vasc. Interven. Radiol 2012;23:3-9).
Dr. Zafar and his colleagues reported that VS outcomes were significantly worse than when the procedures were performed by an IR or IC. IRs had a 32% lower likelihood of ICU use (P less than .001) and a 37% lower likelihood of repeat lower-extremity revascularization or amputation (P less than .001) compared with VS. "Although statistical significance was not reached, both transfusion use and in-hospital mortality were 19% less likely after IRs performed procedures compared with VSs (P = .113 and P =.351, respectively)."
"Vascular surgeons were the only specialists with post-index procedure length of stay exceeding 3 days, significantly longer than observed for other specialties. The adjusted average 1-year costs per index procedure were 9% greater for vascular surgeons than for interventional radiologists ($19,012 vs. $17,640)," they said.
"The reasons for worse outcomes among VSs are not known, but may be related to insufficient training in catheter-based interventions as a result of the extensive time learning and practicing open surgical procedure compared with IRs and ICs, whose focus is catheter interventions," they said. "Medicare data indicate that patients who need lower-extremity endovascular revascularization services experience shorter hospital stays, require less transfusions and ICU services, have lower in-hospital mortality rates, and have much less chance of a subsequent revascularization or amputation within one year if treated by an IR rather than a VS," they concluded.
The study was supported by a grant from the Society of Interventional Radiology and internal funds from the Vascular Disease Research Center. One authors received research funding from Cordis/Johnson&Johnson, and Abbott Vascular and consulting for Microvention/Terumo.
Endovascular lower-extremity revascularization by vascular specialists resulted in more transfusions and ICU use, as well as a longer hospital stay, more repeat revascularization procedures and amputations, and higher costs, compared with the same procedures performed by interventional radiologists, according to the results of a large database analysis of the Medicare 5% Standard Analytical Files.
Dr. Abdul M. Zafar of the Vascular Disease Research Center, Brown University Alpert Medical School, Providence, and his colleagues extracted claims for endovascularlower extremity revascularization (ELER) index procedures using Current Procedural Terminology (CPT) codes in the years 2005 through 2007.
They recorded the unique IDs of the 15,455 patients who underwent these procedures and for whom an inpatient or outpatient claim corresponding to the index procedure claim was available. This comprised all claims for percutaneous angioplasty, atherectomy, and stent implantation of lower extremity arteries.
The self-designated specialty code of the physician performing the procedure was used to determine specialist type: vascular surgeons (VS), interventional radiologists (IR); interventional cardiologists (IC); and "other."
After excluding patients who might have undergone hybrid procedures and those who underwent thrombolysis procedures in the same year as the index procedure, the final sample of 14,608 patients was analyzed for the following outcomes: length of hospital stay; use of intensive care unit (ICU) services; transfusions; in-hospital mortality; and repeat intervention (defined as any ELER, open lower-extremity revascularization, or amputation of the lower extremity).
The researchers built risk-adjusted logistic regression models by using maximum-likelihood estimates to compare various patient outcomes across different specialties, and they used a linear regression model employing ordinary least squares to analyze length of stay. Cost analysis was performed using a linear regression model based on the least-squares approach. All models were adjusted for age, sex, race, admission type (emergency or ambulatory), and other comorbidities using the Elixhauser comorbidities software available from the Agency for Healthcare Research and Quality. Other regression models that included the International Classification of Disease-9 code for disease severity were developed, but were less predictive and had lower R2 values, according to the authors. All logistic regression models passed the goodness-of-fit test at the 99% confidence level. In addition, linear regression models were also found to be satisfactory at the same level, according to the authors.
Within the study, there were 3,565 index procedures done by IRs; 5,489 by ICs; 5,358 by VSs; and 196 performed by "other" specialties (J. Vasc. Interven. Radiol 2012;23:3-9).
Dr. Zafar and his colleagues reported that VS outcomes were significantly worse than when the procedures were performed by an IR or IC. IRs had a 32% lower likelihood of ICU use (P less than .001) and a 37% lower likelihood of repeat lower-extremity revascularization or amputation (P less than .001) compared with VS. "Although statistical significance was not reached, both transfusion use and in-hospital mortality were 19% less likely after IRs performed procedures compared with VSs (P = .113 and P =.351, respectively)."
"Vascular surgeons were the only specialists with post-index procedure length of stay exceeding 3 days, significantly longer than observed for other specialties. The adjusted average 1-year costs per index procedure were 9% greater for vascular surgeons than for interventional radiologists ($19,012 vs. $17,640)," they said.
"The reasons for worse outcomes among VSs are not known, but may be related to insufficient training in catheter-based interventions as a result of the extensive time learning and practicing open surgical procedure compared with IRs and ICs, whose focus is catheter interventions," they said. "Medicare data indicate that patients who need lower-extremity endovascular revascularization services experience shorter hospital stays, require less transfusions and ICU services, have lower in-hospital mortality rates, and have much less chance of a subsequent revascularization or amputation within one year if treated by an IR rather than a VS," they concluded.
The study was supported by a grant from the Society of Interventional Radiology and internal funds from the Vascular Disease Research Center. One authors received research funding from Cordis/Johnson&Johnson, and Abbott Vascular and consulting for Microvention/Terumo.