Western Thoracic Surgical Association (WTSA): Annual Meeting

COLORADO SPRINGS – Using bilateral internal mammary artery grafts provided a significant long-term survival advantage over single mammary artery grafts for coronary artery bypass surgery patients with normal or moderately impaired left ventricular function, according to a large retrospective study with lengthy follow-up.

But when preoperative left ventricular ejection fraction (EF) was less than 30%, the procedure choice made no difference from a survival standpoint.

"BIMA grafting is the operation of choice in patients with a life expectancy beyond 1-2 decades," Dr. David Galbut declared at the annual meeting of the Western Thoracic Surgical Association.

He reported on 4,537 consecutive patients who underwent CABG with internal mammary artery grafting during 1972-1994 at three Florida hospitals. BIMA grafts were performed in 48% of the patients, an exceptionally high BIMA rate. In contrast, the Society of Thoracic Surgeons database shows that, nationally, 4% of patients undergoing CABG receive BIMA grafts. The reason for the 12-fold higher BIMA rate in the Florida study is that Dr. Galbut and his coinvestigators have a decades-long conviction that BIMA offers major clinical advantages.

In the Florida study, 233 patients had an EF below 30%, another 1,256 had an EF of 30%-50%, and 3,048 had a normal EF. In the low EF group, 87 BIMA patients were matched to an equal number of SIMA patients on the basis of 14 preoperative variables. In like manner, propensity scores were used to match 448 BIMA patients in the moderately impaired EF group and 1,137 BIMA patients with a normal EF to similar SIMA patients.

One of the reasons many surgeons are reluctant to use BIMA grafting is a concern that it will result in increased in-hospital complications. This wasn’t the case in the Florida series. Indeed, operative morbidity – including sternal wound infection rates – were similar in matched groups receiving BIMA or SIMA, according to Dr. Galbut of the Aventura (Fla.) Medical Center.

The 20-year survival rate in BIMA patients with moderately impaired EF was 33.1%, significantly better than the 19% survival in matched SIMA patients. In the normal EF group, the 20-year survival rate was 38.1% with BIMA and 35.8% with SIMA.

The general strategy the surgeons followed in BIMA grafting was to run the left internal mammary artery (LIMA) graft to the left anterior descending coronary artery. The LIMA is the dominant vessel in most patients and would therefore be the most durable conduit, Dr. Galbut explained. The right internal mammary artery graft was placed wherever it fit best.

Discussant Dr. Anthony P. Furnary was quick to observe that retrospective studies can’t prove causality, not even when they’re large, painstakingly performed, and feature more than 2 decades of follow-up, as did this one.

He suspects that limitations in the propensity score matching may account for much or all of the long-term survival advantage observed with BIMA grafting in this study. Although patients were extensively matched in terms of 14 preoperative variables, the year of surgery wasn’t among them.

The 22-year study period beginning in 1972 saw the introduction of many modern myocardial protection techniques. If more SIMA patients were operated on in the earlier years of the study, they might well have missed out on the salutary effects of these adjunctive therapies, said Dr. Furnary of the Providence Heart and Vascular Institute in Portland, Ore.

Dr. Galbut declared having no financial conflicts.

COLORADO SPRINGS – Using bilateral internal mammary artery grafts provided a significant long-term survival advantage over single mammary artery grafts for coronary artery bypass surgery patients with normal or moderately impaired left ventricular function, according to a large retrospective study with lengthy follow-up.

But when preoperative left ventricular ejection fraction (EF) was less than 30%, the procedure choice made no difference from a survival standpoint.

"BIMA grafting is the operation of choice in patients with a life expectancy beyond 1-2 decades," Dr. David Galbut declared at the annual meeting of the Western Thoracic Surgical Association.

He reported on 4,537 consecutive patients who underwent CABG with internal mammary artery grafting during 1972-1994 at three Florida hospitals. BIMA grafts were performed in 48% of the patients, an exceptionally high BIMA rate. In contrast, the Society of Thoracic Surgeons database shows that, nationally, 4% of patients undergoing CABG receive BIMA grafts. The reason for the 12-fold higher BIMA rate in the Florida study is that Dr. Galbut and his coinvestigators have a decades-long conviction that BIMA offers major clinical advantages.

In the Florida study, 233 patients had an EF below 30%, another 1,256 had an EF of 30%-50%, and 3,048 had a normal EF. In the low EF group, 87 BIMA patients were matched to an equal number of SIMA patients on the basis of 14 preoperative variables. In like manner, propensity scores were used to match 448 BIMA patients in the moderately impaired EF group and 1,137 BIMA patients with a normal EF to similar SIMA patients.

One of the reasons many surgeons are reluctant to use BIMA grafting is a concern that it will result in increased in-hospital complications. This wasn’t the case in the Florida series. Indeed, operative morbidity – including sternal wound infection rates – were similar in matched groups receiving BIMA or SIMA, according to Dr. Galbut of the Aventura (Fla.) Medical Center.

The 20-year survival rate in BIMA patients with moderately impaired EF was 33.1%, significantly better than the 19% survival in matched SIMA patients. In the normal EF group, the 20-year survival rate was 38.1% with BIMA and 35.8% with SIMA.

The general strategy the surgeons followed in BIMA grafting was to run the left internal mammary artery (LIMA) graft to the left anterior descending coronary artery. The LIMA is the dominant vessel in most patients and would therefore be the most durable conduit, Dr. Galbut explained. The right internal mammary artery graft was placed wherever it fit best.

Discussant Dr. Anthony P. Furnary was quick to observe that retrospective studies can’t prove causality, not even when they’re large, painstakingly performed, and feature more than 2 decades of follow-up, as did this one.

He suspects that limitations in the propensity score matching may account for much or all of the long-term survival advantage observed with BIMA grafting in this study. Although patients were extensively matched in terms of 14 preoperative variables, the year of surgery wasn’t among them.

The 22-year study period beginning in 1972 saw the introduction of many modern myocardial protection techniques. If more SIMA patients were operated on in the earlier years of the study, they might well have missed out on the salutary effects of these adjunctive therapies, said Dr. Furnary of the Providence Heart and Vascular Institute in Portland, Ore.

Dr. Galbut declared having no financial conflicts.

COLORADO SPRINGS – Using bilateral internal mammary artery grafts provided a significant long-term survival advantage over single mammary artery grafts for coronary artery bypass surgery patients with normal or moderately impaired left ventricular function, according to a large retrospective study with lengthy follow-up.

But when preoperative left ventricular ejection fraction (EF) was less than 30%, the procedure choice made no difference from a survival standpoint.

"BIMA grafting is the operation of choice in patients with a life expectancy beyond 1-2 decades," Dr. David Galbut declared at the annual meeting of the Western Thoracic Surgical Association.

He reported on 4,537 consecutive patients who underwent CABG with internal mammary artery grafting during 1972-1994 at three Florida hospitals. BIMA grafts were performed in 48% of the patients, an exceptionally high BIMA rate. In contrast, the Society of Thoracic Surgeons database shows that, nationally, 4% of patients undergoing CABG receive BIMA grafts. The reason for the 12-fold higher BIMA rate in the Florida study is that Dr. Galbut and his coinvestigators have a decades-long conviction that BIMA offers major clinical advantages.

In the Florida study, 233 patients had an EF below 30%, another 1,256 had an EF of 30%-50%, and 3,048 had a normal EF. In the low EF group, 87 BIMA patients were matched to an equal number of SIMA patients on the basis of 14 preoperative variables. In like manner, propensity scores were used to match 448 BIMA patients in the moderately impaired EF group and 1,137 BIMA patients with a normal EF to similar SIMA patients.

One of the reasons many surgeons are reluctant to use BIMA grafting is a concern that it will result in increased in-hospital complications. This wasn’t the case in the Florida series. Indeed, operative morbidity – including sternal wound infection rates – were similar in matched groups receiving BIMA or SIMA, according to Dr. Galbut of the Aventura (Fla.) Medical Center.

The 20-year survival rate in BIMA patients with moderately impaired EF was 33.1%, significantly better than the 19% survival in matched SIMA patients. In the normal EF group, the 20-year survival rate was 38.1% with BIMA and 35.8% with SIMA.

The general strategy the surgeons followed in BIMA grafting was to run the left internal mammary artery (LIMA) graft to the left anterior descending coronary artery. The LIMA is the dominant vessel in most patients and would therefore be the most durable conduit, Dr. Galbut explained. The right internal mammary artery graft was placed wherever it fit best.

Discussant Dr. Anthony P. Furnary was quick to observe that retrospective studies can’t prove causality, not even when they’re large, painstakingly performed, and feature more than 2 decades of follow-up, as did this one.

He suspects that limitations in the propensity score matching may account for much or all of the long-term survival advantage observed with BIMA grafting in this study. Although patients were extensively matched in terms of 14 preoperative variables, the year of surgery wasn’t among them.

The 22-year study period beginning in 1972 saw the introduction of many modern myocardial protection techniques. If more SIMA patients were operated on in the earlier years of the study, they might well have missed out on the salutary effects of these adjunctive therapies, said Dr. Furnary of the Providence Heart and Vascular Institute in Portland, Ore.

Dr. Galbut declared having no financial conflicts.

COLORADO SPRINGS – A disproportionate share of donor lungs goes to local, low-priority recipients, according to an analysis of data from the United Network for Organ Sharing.

The current lung allocation system results in a high proportion of donor organs being distributed to low-priority candidates who often receive little survival benefit from their transplant. Meanwhile, higher-priority candidates who might derive more benefit from transplantation continue to die at high rates on the waiting list, Dr. Alexander Iribarne said at the annual meeting of the Western Thoracic Surgical Association.

He presented an argument for the sharing of donor lungs over a broader geographical range in the United States. The analysis involved all 7,171 lung transplants done in the United States between May 2005 and the end of 2010.

May 2005 was chosen as the starting date because that’s when the Lung Allocation Score (LAS) was introduced as a measure for allocating organs on the basis of medical urgency rather than waiting time. An LAS score lower than 50 defines a transplant candidate as low priority, a score of 50-75 is considered intermediate, and an LAS greater than 75 is high priority.

Among the 5,544 transplants that were performed in low-priority recipients, 54% of the donor organs were allocated locally, 17% regionally, and 29% nationally. In contrast, 40% of the 1,016 transplants in recipients with an LAS of 50-75 at the time of surgery were allocated locally, as were 33% of the donor organs used in patients with an LAS greater than 75.

What’s happening is that when an organ becomes available in one of the less-populated local donor service areas, there’s a lower likelihood that a suitable higher-priority candidate will be in place than in a more populous donor service area, said Dr. Iribarne of Columbia University, New York.

As a result, the organ often goes to a local patient with an LAS lower than 50.

The UNOS data showed that in donor service areas having a population of fewer than 6.1 million, nearly 75% of locally allocated donor lungs went to patients with an LAS lower than 50. In local donor service areas with a population in excess of 10.3 million people, a greater percentage of lungs are allocated to higher-priority recipients, according to Dr. Iribarne.

The next step in this research should be to determine whether organ sharing across broader geographical areas results in higher rates of lung allocation to higher-priority candidates and improved survival for those on the waiting list, he added.

Dr. Iribarne declared having no financial conflicts.

COLORADO SPRINGS – A disproportionate share of donor lungs goes to local, low-priority recipients, according to an analysis of data from the United Network for Organ Sharing.

The current lung allocation system results in a high proportion of donor organs being distributed to low-priority candidates who often receive little survival benefit from their transplant. Meanwhile, higher-priority candidates who might derive more benefit from transplantation continue to die at high rates on the waiting list, Dr. Alexander Iribarne said at the annual meeting of the Western Thoracic Surgical Association.

He presented an argument for the sharing of donor lungs over a broader geographical range in the United States. The analysis involved all 7,171 lung transplants done in the United States between May 2005 and the end of 2010.

May 2005 was chosen as the starting date because that’s when the Lung Allocation Score (LAS) was introduced as a measure for allocating organs on the basis of medical urgency rather than waiting time. An LAS score lower than 50 defines a transplant candidate as low priority, a score of 50-75 is considered intermediate, and an LAS greater than 75 is high priority.

Among the 5,544 transplants that were performed in low-priority recipients, 54% of the donor organs were allocated locally, 17% regionally, and 29% nationally. In contrast, 40% of the 1,016 transplants in recipients with an LAS of 50-75 at the time of surgery were allocated locally, as were 33% of the donor organs used in patients with an LAS greater than 75.

What’s happening is that when an organ becomes available in one of the less-populated local donor service areas, there’s a lower likelihood that a suitable higher-priority candidate will be in place than in a more populous donor service area, said Dr. Iribarne of Columbia University, New York.

As a result, the organ often goes to a local patient with an LAS lower than 50.

The UNOS data showed that in donor service areas having a population of fewer than 6.1 million, nearly 75% of locally allocated donor lungs went to patients with an LAS lower than 50. In local donor service areas with a population in excess of 10.3 million people, a greater percentage of lungs are allocated to higher-priority recipients, according to Dr. Iribarne.

The next step in this research should be to determine whether organ sharing across broader geographical areas results in higher rates of lung allocation to higher-priority candidates and improved survival for those on the waiting list, he added.

Dr. Iribarne declared having no financial conflicts.

COLORADO SPRINGS – A disproportionate share of donor lungs goes to local, low-priority recipients, according to an analysis of data from the United Network for Organ Sharing.

The current lung allocation system results in a high proportion of donor organs being distributed to low-priority candidates who often receive little survival benefit from their transplant. Meanwhile, higher-priority candidates who might derive more benefit from transplantation continue to die at high rates on the waiting list, Dr. Alexander Iribarne said at the annual meeting of the Western Thoracic Surgical Association.

He presented an argument for the sharing of donor lungs over a broader geographical range in the United States. The analysis involved all 7,171 lung transplants done in the United States between May 2005 and the end of 2010.

May 2005 was chosen as the starting date because that’s when the Lung Allocation Score (LAS) was introduced as a measure for allocating organs on the basis of medical urgency rather than waiting time. An LAS score lower than 50 defines a transplant candidate as low priority, a score of 50-75 is considered intermediate, and an LAS greater than 75 is high priority.

Among the 5,544 transplants that were performed in low-priority recipients, 54% of the donor organs were allocated locally, 17% regionally, and 29% nationally. In contrast, 40% of the 1,016 transplants in recipients with an LAS of 50-75 at the time of surgery were allocated locally, as were 33% of the donor organs used in patients with an LAS greater than 75.

What’s happening is that when an organ becomes available in one of the less-populated local donor service areas, there’s a lower likelihood that a suitable higher-priority candidate will be in place than in a more populous donor service area, said Dr. Iribarne of Columbia University, New York.

As a result, the organ often goes to a local patient with an LAS lower than 50.

The UNOS data showed that in donor service areas having a population of fewer than 6.1 million, nearly 75% of locally allocated donor lungs went to patients with an LAS lower than 50. In local donor service areas with a population in excess of 10.3 million people, a greater percentage of lungs are allocated to higher-priority recipients, according to Dr. Iribarne.

The next step in this research should be to determine whether organ sharing across broader geographical areas results in higher rates of lung allocation to higher-priority candidates and improved survival for those on the waiting list, he added.

Dr. Iribarne declared having no financial conflicts.

COLORADO SPRINGS – Physicians at Emory University Hospital now have a firmer grip on what they’re dealing with in reducing 30-day readmissions.

The knowledge that there are three major risk factors – preoperative failure to thrive, an initial length of stay greater than 10 days, and Hispanic ethnicity – for readmission following pediatric congenital heart disease surgery is a tool for improvement, said Dr. Brian E. Kogon of Emory University, Atlanta.

"These data are obviously our data and are very specific to our hospital, our location, and our patient population. I would think that it’s going to be very different throughout the country based on whether you’re at an academic center or private center, urban versus rural setting, and even in adult cardiac, general thoracic, and pediatric practices," Dr. Kogon said in presenting the study findings at the annual meeting of the Western Thoracic Surgical Association.

The important thing is for physicians and surgeons to analyze their own hospital’s readmission experience, identify the risk factors, and then address the potentially modifiable ones in an effort to drive that readmission rate down, added Dr. Kogon, director of the congenital cardiac surgery fellowship program at Emory.

Studying Readmissions Following Pediatric Surgery

Readmissions within 30 days are increasingly viewed by third-party payers as preventable complications warranting stiff payment penalties. The focus thus far has been on the adult world, but at some point pediatric care will come under scrutiny as well. This realization led Dr. Kogon and his coworkers to analyze their institutional experience via a retrospective cohort study.

During 2002-2009, the annual 30-day readmission rates following pediatric surgery for congenital heart disease were 5.9%-10.4%, with a median of 8.7%. Those rates are relatively low; other centers typically report readmission rates of 10%-20%, he noted.

In 2009, 685 patients were discharged after pediatric congenital heart disease surgery; 70 of them had 74 readmissions. Among the key findings: only 15% of readmissions were for cardiac reasons. Indeed, the top three reasons for readmission were pleural or pericardial effusions, accounting for 26% of all readmissions; gastrointestinal problems, 24%; and infection, 19%.

Readmissions were costly. A total of 69% of patients were readmitted to a ward, 31% to the ICU. Upon readmission these patients spent a total of 653 additional days – almost 22 months – in the hospital.

The investigators scrutinized numerous potential demographic, preoperative, operative, and postoperative risk factors for readmission. Only three proved significant in a multivariate analysis: an initial length of stay greater than 10 days was associated with a 4.4-fold increased risk of readmission; a preoperative diagnosis of failure to thrive was associated with a 2.7-fold risk; and Hispanic ethnicity was associated with a 1.87-fold increased risk.

These readmissions occurred despite an intense discharge process and close follow-up. All families at the pediatric heart surgery unit attend a discharge class and a CPR training class. A pharmacist is on hand at the discharge class to review medications. Shunt recipients and newborns receive additional education. All case-management issues, such as formula supplies and home health equipment, are resolved before discharge. Patients meet with a cardiothoracic surgeon during their first week out of hospital, a cardiologist the second week, and thereafter with their primary care provider.

The median time to the first scheduled outpatient appointment was 4.5 days postdischarge. The median time to readmission was 8 days. Thirty-one percent of patients were readmitted prior to their first clinic appointment, 10% directly from the clinic, and 50% after their first clinic visit. The rest were readmitted after being no-shows for their clinic visit.

Dr. Kogon said he suspects that Hispanics were at increased risk for readmission because of educational and language barriers. Although a Spanish-language interpreter is present at the discharge class as needed, Dr. Kogon and his colleagues have observed that many Hispanic families nonetheless return unclear about medication and feeding regimens.

"I think there’s still a gap in our education of those patients," he said.

Discussant Dr. David R. Clarke said it might be argued that if a hospital doesn’t have a certain number of readmissions, then patients are being kept in the hospital too long.

"On a practical level, how much do we spend during the initial admission to ensure no readmissions? Do we automatically keep patients identified as high risk, such as Hispanics, 2, 3, or 4 extra days to minimize their readmission rate? And even if we do that and other things, is it really possible to prevent readmissions?" wondered Dr. Clarke of Children’s Hospital Colorado, Denver.

Dr. Kogon replied that this is the key question his colleagues raised when he shared the study findings. The group has decided to modify the discharge process for their high-risk patients, keeping them in the hospital a day or so longer while continuing to collect data in order to see if this pays off in fewer readmissions.

Dr. Kogon declared having no financial conflicts.

COLORADO SPRINGS – Physicians at Emory University Hospital now have a firmer grip on what they’re dealing with in reducing 30-day readmissions.

The knowledge that there are three major risk factors – preoperative failure to thrive, an initial length of stay greater than 10 days, and Hispanic ethnicity – for readmission following pediatric congenital heart disease surgery is a tool for improvement, said Dr. Brian E. Kogon of Emory University, Atlanta.

"These data are obviously our data and are very specific to our hospital, our location, and our patient population. I would think that it’s going to be very different throughout the country based on whether you’re at an academic center or private center, urban versus rural setting, and even in adult cardiac, general thoracic, and pediatric practices," Dr. Kogon said in presenting the study findings at the annual meeting of the Western Thoracic Surgical Association.

The important thing is for physicians and surgeons to analyze their own hospital’s readmission experience, identify the risk factors, and then address the potentially modifiable ones in an effort to drive that readmission rate down, added Dr. Kogon, director of the congenital cardiac surgery fellowship program at Emory.

Studying Readmissions Following Pediatric Surgery

Readmissions within 30 days are increasingly viewed by third-party payers as preventable complications warranting stiff payment penalties. The focus thus far has been on the adult world, but at some point pediatric care will come under scrutiny as well. This realization led Dr. Kogon and his coworkers to analyze their institutional experience via a retrospective cohort study.

During 2002-2009, the annual 30-day readmission rates following pediatric surgery for congenital heart disease were 5.9%-10.4%, with a median of 8.7%. Those rates are relatively low; other centers typically report readmission rates of 10%-20%, he noted.

In 2009, 685 patients were discharged after pediatric congenital heart disease surgery; 70 of them had 74 readmissions. Among the key findings: only 15% of readmissions were for cardiac reasons. Indeed, the top three reasons for readmission were pleural or pericardial effusions, accounting for 26% of all readmissions; gastrointestinal problems, 24%; and infection, 19%.

Readmissions were costly. A total of 69% of patients were readmitted to a ward, 31% to the ICU. Upon readmission these patients spent a total of 653 additional days – almost 22 months – in the hospital.

The investigators scrutinized numerous potential demographic, preoperative, operative, and postoperative risk factors for readmission. Only three proved significant in a multivariate analysis: an initial length of stay greater than 10 days was associated with a 4.4-fold increased risk of readmission; a preoperative diagnosis of failure to thrive was associated with a 2.7-fold risk; and Hispanic ethnicity was associated with a 1.87-fold increased risk.

These readmissions occurred despite an intense discharge process and close follow-up. All families at the pediatric heart surgery unit attend a discharge class and a CPR training class. A pharmacist is on hand at the discharge class to review medications. Shunt recipients and newborns receive additional education. All case-management issues, such as formula supplies and home health equipment, are resolved before discharge. Patients meet with a cardiothoracic surgeon during their first week out of hospital, a cardiologist the second week, and thereafter with their primary care provider.

The median time to the first scheduled outpatient appointment was 4.5 days postdischarge. The median time to readmission was 8 days. Thirty-one percent of patients were readmitted prior to their first clinic appointment, 10% directly from the clinic, and 50% after their first clinic visit. The rest were readmitted after being no-shows for their clinic visit.

Dr. Kogon said he suspects that Hispanics were at increased risk for readmission because of educational and language barriers. Although a Spanish-language interpreter is present at the discharge class as needed, Dr. Kogon and his colleagues have observed that many Hispanic families nonetheless return unclear about medication and feeding regimens.

"I think there’s still a gap in our education of those patients," he said.

Discussant Dr. David R. Clarke said it might be argued that if a hospital doesn’t have a certain number of readmissions, then patients are being kept in the hospital too long.

"On a practical level, how much do we spend during the initial admission to ensure no readmissions? Do we automatically keep patients identified as high risk, such as Hispanics, 2, 3, or 4 extra days to minimize their readmission rate? And even if we do that and other things, is it really possible to prevent readmissions?" wondered Dr. Clarke of Children’s Hospital Colorado, Denver.

Dr. Kogon replied that this is the key question his colleagues raised when he shared the study findings. The group has decided to modify the discharge process for their high-risk patients, keeping them in the hospital a day or so longer while continuing to collect data in order to see if this pays off in fewer readmissions.

Dr. Kogon declared having no financial conflicts.

COLORADO SPRINGS – Physicians at Emory University Hospital now have a firmer grip on what they’re dealing with in reducing 30-day readmissions.

The knowledge that there are three major risk factors – preoperative failure to thrive, an initial length of stay greater than 10 days, and Hispanic ethnicity – for readmission following pediatric congenital heart disease surgery is a tool for improvement, said Dr. Brian E. Kogon of Emory University, Atlanta.

"These data are obviously our data and are very specific to our hospital, our location, and our patient population. I would think that it’s going to be very different throughout the country based on whether you’re at an academic center or private center, urban versus rural setting, and even in adult cardiac, general thoracic, and pediatric practices," Dr. Kogon said in presenting the study findings at the annual meeting of the Western Thoracic Surgical Association.

The important thing is for physicians and surgeons to analyze their own hospital’s readmission experience, identify the risk factors, and then address the potentially modifiable ones in an effort to drive that readmission rate down, added Dr. Kogon, director of the congenital cardiac surgery fellowship program at Emory.

Studying Readmissions Following Pediatric Surgery

Readmissions within 30 days are increasingly viewed by third-party payers as preventable complications warranting stiff payment penalties. The focus thus far has been on the adult world, but at some point pediatric care will come under scrutiny as well. This realization led Dr. Kogon and his coworkers to analyze their institutional experience via a retrospective cohort study.

During 2002-2009, the annual 30-day readmission rates following pediatric surgery for congenital heart disease were 5.9%-10.4%, with a median of 8.7%. Those rates are relatively low; other centers typically report readmission rates of 10%-20%, he noted.

In 2009, 685 patients were discharged after pediatric congenital heart disease surgery; 70 of them had 74 readmissions. Among the key findings: only 15% of readmissions were for cardiac reasons. Indeed, the top three reasons for readmission were pleural or pericardial effusions, accounting for 26% of all readmissions; gastrointestinal problems, 24%; and infection, 19%.

Readmissions were costly. A total of 69% of patients were readmitted to a ward, 31% to the ICU. Upon readmission these patients spent a total of 653 additional days – almost 22 months – in the hospital.

The investigators scrutinized numerous potential demographic, preoperative, operative, and postoperative risk factors for readmission. Only three proved significant in a multivariate analysis: an initial length of stay greater than 10 days was associated with a 4.4-fold increased risk of readmission; a preoperative diagnosis of failure to thrive was associated with a 2.7-fold risk; and Hispanic ethnicity was associated with a 1.87-fold increased risk.

These readmissions occurred despite an intense discharge process and close follow-up. All families at the pediatric heart surgery unit attend a discharge class and a CPR training class. A pharmacist is on hand at the discharge class to review medications. Shunt recipients and newborns receive additional education. All case-management issues, such as formula supplies and home health equipment, are resolved before discharge. Patients meet with a cardiothoracic surgeon during their first week out of hospital, a cardiologist the second week, and thereafter with their primary care provider.

The median time to the first scheduled outpatient appointment was 4.5 days postdischarge. The median time to readmission was 8 days. Thirty-one percent of patients were readmitted prior to their first clinic appointment, 10% directly from the clinic, and 50% after their first clinic visit. The rest were readmitted after being no-shows for their clinic visit.

Dr. Kogon said he suspects that Hispanics were at increased risk for readmission because of educational and language barriers. Although a Spanish-language interpreter is present at the discharge class as needed, Dr. Kogon and his colleagues have observed that many Hispanic families nonetheless return unclear about medication and feeding regimens.

"I think there’s still a gap in our education of those patients," he said.

Discussant Dr. David R. Clarke said it might be argued that if a hospital doesn’t have a certain number of readmissions, then patients are being kept in the hospital too long.

"On a practical level, how much do we spend during the initial admission to ensure no readmissions? Do we automatically keep patients identified as high risk, such as Hispanics, 2, 3, or 4 extra days to minimize their readmission rate? And even if we do that and other things, is it really possible to prevent readmissions?" wondered Dr. Clarke of Children’s Hospital Colorado, Denver.

Dr. Kogon replied that this is the key question his colleagues raised when he shared the study findings. The group has decided to modify the discharge process for their high-risk patients, keeping them in the hospital a day or so longer while continuing to collect data in order to see if this pays off in fewer readmissions.

Dr. Kogon declared having no financial conflicts.