User login
NEW ORLEANS – The 3M Potentially Preventable Readmissions algorithm is not ready for prime time as a tool to assess preventable pediatric readmissions, preliminary data from an analysis of 1.7 million admissions suggests.
The analysis found that only about 20% of all pediatric readmissions are potentially preventable, whereas the PPR algorithm put that number at about 40%, Dr. James Gay said at the Pediatric Hospital Medicine 2013 meeting.
The results are concerning because the PPR algorithm has been used by several state Medicaid programs as the basis for levying financial penalties on hospitals with high readmission rates, including some children’s hospitals.
"We’re not at the point where children’s hospitals should be penalized by any system that’s judging readmission preventable, because we just don’t have the tools for that to be fair," he said. "I’m hoping 3M will use this [data] to apply overlays to their algorithm – socioeconomic overlays, preventability scoring factors – but this will take a lot more work to decide what’s really preventable or not."
The 3M algorithm was designed for use in patients of all ages and uses administrative data to identify "potentially preventable" readmissions based on the relationship between All Patient Refined Diagnosis Related Groups (APR-DRGs) of the index admission and readmissions. It is thought to have a potential advantage over an "all-cause" approach because it has many predefined exclusions such as APR-DRGs for chemotherapy and cystic fibrosis.
The problem is the algorithm may not be sensitive to pediatric readmissions, which differ from those in adults in diagnoses, rates, and preventability, said Dr. Gay, with the Monroe Carell Jr. Children’s Hospital at Vanderbilt, in Nashville.
The investigators analyzed data from 1,719,617 hospitalizations for 1,496,470 patients admitted at 58 Children’s Hospital Association hospitals from 2009 to 2011. They compared the PPR algorithm readmission rates with all-cause readmission rates at 7, 15, and 30 days, determined which diagnoses accounted for the largest volume and cost of readmission by both methods, and accessed the financial impact of the readmissions.
At 7, 15, and 30 days, the PPR algorithm readmission rates were 2.5%, 4.1%, and 6.3% or almost exactly half of the all-cause readmission rates of 5%, 8.7%, and 13.3%, respectively, Dr. Gay said.
The same was true for the total number of PPR admissions at 11,898, 19,724, and 30,226, respectively, versus all-cause readmissions of 28,755, 49,982, and 76,245, respectively.
For each time period, the overall percentage of potentially preventable readmissions with the PPR algorithm hovered around 40%, with very little variability, Dr. Gay said at the meeting, which was sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association.
"This runs somewhat contrary to the common notion that a readmission is more likely to be preventable the earlier it occurs after discharge, but the PPR system does not pick up on that," he observed.
Notably, the PPR algorithm considered at least 80% of readmissions preventable for six APR-DRGs: sickle cell crisis (81%), asthma (81%), bronchiolitis (83%), ventricular shunt procedures (86%), connective tissue disorders (87%), and appendectomy (100%).
"Parenthetically, there are a number of pediatricians that might argue about 80% of bronchiolitis readmissions being preventable," he quipped.
Finally, the total annual cost for admissions in the study was $11.6 billion, of which $1.7 billion (14.7%) was related to all-cause readmissions and $533 million (4.6%) to PPR algorithm readmissions.
If the data are extrapolated, the estimated nationwide annual cost for pediatric hospitalizations is $33.6 billion, with a potential savings of $1.5 billion per year (4.6%) if all PPR readmissions could be prevented. Based on a crude back-of-the-envelope calculation, the $1.5 billion pales in comparison to the $25 billion projected to be saved annually in preventable Medicare readmissions, Dr. Gay observed.
"The financial yield in preventing PPR readmissions in pediatric populations is relatively small, and efforts should continue to assure the highest-quality inpatient care and discharge planning," he said.
"Further research is needed to validate the applicability and appropriateness of PPR inclusions and exclusions in pediatric populations before it can be recommended for such consequential purposes as financial penalties to children’s hospitals."
During a discussion of the findings, an attendee asked whether this type of research was worth burning up so many bright minds, given that the data have repeatedly shown that pediatric readmission rates are so low. Dr. Gay responded, saying that, quite frankly, the work is necessary to avoid financial penalties and to drive home the point that quality improvement measures focused on pediatric readmissions are "bogus."
In a statement responding to Dr. Gay’s analysis, 3M Health Information Systems’ senior vice president of clinical and economic research, Richard Averill, said, "In payment applications, the number of PPRs for each provider is compared on a risk-adjusted basis to identify those providers with an ‘excess’ number of readmissions that are potentially preventable. Any payment adjustment for readmission is based only on the extent of excess readmissions.
"If all providers had the same PPR rate, no provider would have any excess PPRs and there would be no payment adjustment for readmission. Unfortunately, Dr. Gay’s comments omit any mention of how PPRs are actually being used and that they are not intended to provide a direct measure of the number of preventable readmissions."
Mr. Averill noted that Dr. Gay’s analysis observes that "half of readmissions detected by an ‘all-cause’ method would not be considered potentially preventable using PPRs. An examination of excess PPR rates across children’s hospitals can be an important quality improvement tool," he said.
In a separate interview, Dr. Shawn Ralston, chief of the pediatrics section, Dartmouth-Hitchcock Medical Center, Lebanon, N.H., agreed with Dr. Gay and said there are more pressing issues in pediatric clinical quality that deserve attention.
"I want meaningful quality measures developed by people with pediatric expertise and not things extrapolated from adult patients," Dr. Ralston added.
Dr. Gay reported serving as the medical consultant to the Children’s Hospital Association.
NEW ORLEANS – The 3M Potentially Preventable Readmissions algorithm is not ready for prime time as a tool to assess preventable pediatric readmissions, preliminary data from an analysis of 1.7 million admissions suggests.
The analysis found that only about 20% of all pediatric readmissions are potentially preventable, whereas the PPR algorithm put that number at about 40%, Dr. James Gay said at the Pediatric Hospital Medicine 2013 meeting.
The results are concerning because the PPR algorithm has been used by several state Medicaid programs as the basis for levying financial penalties on hospitals with high readmission rates, including some children’s hospitals.
"We’re not at the point where children’s hospitals should be penalized by any system that’s judging readmission preventable, because we just don’t have the tools for that to be fair," he said. "I’m hoping 3M will use this [data] to apply overlays to their algorithm – socioeconomic overlays, preventability scoring factors – but this will take a lot more work to decide what’s really preventable or not."
The 3M algorithm was designed for use in patients of all ages and uses administrative data to identify "potentially preventable" readmissions based on the relationship between All Patient Refined Diagnosis Related Groups (APR-DRGs) of the index admission and readmissions. It is thought to have a potential advantage over an "all-cause" approach because it has many predefined exclusions such as APR-DRGs for chemotherapy and cystic fibrosis.
The problem is the algorithm may not be sensitive to pediatric readmissions, which differ from those in adults in diagnoses, rates, and preventability, said Dr. Gay, with the Monroe Carell Jr. Children’s Hospital at Vanderbilt, in Nashville.
The investigators analyzed data from 1,719,617 hospitalizations for 1,496,470 patients admitted at 58 Children’s Hospital Association hospitals from 2009 to 2011. They compared the PPR algorithm readmission rates with all-cause readmission rates at 7, 15, and 30 days, determined which diagnoses accounted for the largest volume and cost of readmission by both methods, and accessed the financial impact of the readmissions.
At 7, 15, and 30 days, the PPR algorithm readmission rates were 2.5%, 4.1%, and 6.3% or almost exactly half of the all-cause readmission rates of 5%, 8.7%, and 13.3%, respectively, Dr. Gay said.
The same was true for the total number of PPR admissions at 11,898, 19,724, and 30,226, respectively, versus all-cause readmissions of 28,755, 49,982, and 76,245, respectively.
For each time period, the overall percentage of potentially preventable readmissions with the PPR algorithm hovered around 40%, with very little variability, Dr. Gay said at the meeting, which was sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association.
"This runs somewhat contrary to the common notion that a readmission is more likely to be preventable the earlier it occurs after discharge, but the PPR system does not pick up on that," he observed.
Notably, the PPR algorithm considered at least 80% of readmissions preventable for six APR-DRGs: sickle cell crisis (81%), asthma (81%), bronchiolitis (83%), ventricular shunt procedures (86%), connective tissue disorders (87%), and appendectomy (100%).
"Parenthetically, there are a number of pediatricians that might argue about 80% of bronchiolitis readmissions being preventable," he quipped.
Finally, the total annual cost for admissions in the study was $11.6 billion, of which $1.7 billion (14.7%) was related to all-cause readmissions and $533 million (4.6%) to PPR algorithm readmissions.
If the data are extrapolated, the estimated nationwide annual cost for pediatric hospitalizations is $33.6 billion, with a potential savings of $1.5 billion per year (4.6%) if all PPR readmissions could be prevented. Based on a crude back-of-the-envelope calculation, the $1.5 billion pales in comparison to the $25 billion projected to be saved annually in preventable Medicare readmissions, Dr. Gay observed.
"The financial yield in preventing PPR readmissions in pediatric populations is relatively small, and efforts should continue to assure the highest-quality inpatient care and discharge planning," he said.
"Further research is needed to validate the applicability and appropriateness of PPR inclusions and exclusions in pediatric populations before it can be recommended for such consequential purposes as financial penalties to children’s hospitals."
During a discussion of the findings, an attendee asked whether this type of research was worth burning up so many bright minds, given that the data have repeatedly shown that pediatric readmission rates are so low. Dr. Gay responded, saying that, quite frankly, the work is necessary to avoid financial penalties and to drive home the point that quality improvement measures focused on pediatric readmissions are "bogus."
In a statement responding to Dr. Gay’s analysis, 3M Health Information Systems’ senior vice president of clinical and economic research, Richard Averill, said, "In payment applications, the number of PPRs for each provider is compared on a risk-adjusted basis to identify those providers with an ‘excess’ number of readmissions that are potentially preventable. Any payment adjustment for readmission is based only on the extent of excess readmissions.
"If all providers had the same PPR rate, no provider would have any excess PPRs and there would be no payment adjustment for readmission. Unfortunately, Dr. Gay’s comments omit any mention of how PPRs are actually being used and that they are not intended to provide a direct measure of the number of preventable readmissions."
Mr. Averill noted that Dr. Gay’s analysis observes that "half of readmissions detected by an ‘all-cause’ method would not be considered potentially preventable using PPRs. An examination of excess PPR rates across children’s hospitals can be an important quality improvement tool," he said.
In a separate interview, Dr. Shawn Ralston, chief of the pediatrics section, Dartmouth-Hitchcock Medical Center, Lebanon, N.H., agreed with Dr. Gay and said there are more pressing issues in pediatric clinical quality that deserve attention.
"I want meaningful quality measures developed by people with pediatric expertise and not things extrapolated from adult patients," Dr. Ralston added.
Dr. Gay reported serving as the medical consultant to the Children’s Hospital Association.
NEW ORLEANS – The 3M Potentially Preventable Readmissions algorithm is not ready for prime time as a tool to assess preventable pediatric readmissions, preliminary data from an analysis of 1.7 million admissions suggests.
The analysis found that only about 20% of all pediatric readmissions are potentially preventable, whereas the PPR algorithm put that number at about 40%, Dr. James Gay said at the Pediatric Hospital Medicine 2013 meeting.
The results are concerning because the PPR algorithm has been used by several state Medicaid programs as the basis for levying financial penalties on hospitals with high readmission rates, including some children’s hospitals.
"We’re not at the point where children’s hospitals should be penalized by any system that’s judging readmission preventable, because we just don’t have the tools for that to be fair," he said. "I’m hoping 3M will use this [data] to apply overlays to their algorithm – socioeconomic overlays, preventability scoring factors – but this will take a lot more work to decide what’s really preventable or not."
The 3M algorithm was designed for use in patients of all ages and uses administrative data to identify "potentially preventable" readmissions based on the relationship between All Patient Refined Diagnosis Related Groups (APR-DRGs) of the index admission and readmissions. It is thought to have a potential advantage over an "all-cause" approach because it has many predefined exclusions such as APR-DRGs for chemotherapy and cystic fibrosis.
The problem is the algorithm may not be sensitive to pediatric readmissions, which differ from those in adults in diagnoses, rates, and preventability, said Dr. Gay, with the Monroe Carell Jr. Children’s Hospital at Vanderbilt, in Nashville.
The investigators analyzed data from 1,719,617 hospitalizations for 1,496,470 patients admitted at 58 Children’s Hospital Association hospitals from 2009 to 2011. They compared the PPR algorithm readmission rates with all-cause readmission rates at 7, 15, and 30 days, determined which diagnoses accounted for the largest volume and cost of readmission by both methods, and accessed the financial impact of the readmissions.
At 7, 15, and 30 days, the PPR algorithm readmission rates were 2.5%, 4.1%, and 6.3% or almost exactly half of the all-cause readmission rates of 5%, 8.7%, and 13.3%, respectively, Dr. Gay said.
The same was true for the total number of PPR admissions at 11,898, 19,724, and 30,226, respectively, versus all-cause readmissions of 28,755, 49,982, and 76,245, respectively.
For each time period, the overall percentage of potentially preventable readmissions with the PPR algorithm hovered around 40%, with very little variability, Dr. Gay said at the meeting, which was sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association.
"This runs somewhat contrary to the common notion that a readmission is more likely to be preventable the earlier it occurs after discharge, but the PPR system does not pick up on that," he observed.
Notably, the PPR algorithm considered at least 80% of readmissions preventable for six APR-DRGs: sickle cell crisis (81%), asthma (81%), bronchiolitis (83%), ventricular shunt procedures (86%), connective tissue disorders (87%), and appendectomy (100%).
"Parenthetically, there are a number of pediatricians that might argue about 80% of bronchiolitis readmissions being preventable," he quipped.
Finally, the total annual cost for admissions in the study was $11.6 billion, of which $1.7 billion (14.7%) was related to all-cause readmissions and $533 million (4.6%) to PPR algorithm readmissions.
If the data are extrapolated, the estimated nationwide annual cost for pediatric hospitalizations is $33.6 billion, with a potential savings of $1.5 billion per year (4.6%) if all PPR readmissions could be prevented. Based on a crude back-of-the-envelope calculation, the $1.5 billion pales in comparison to the $25 billion projected to be saved annually in preventable Medicare readmissions, Dr. Gay observed.
"The financial yield in preventing PPR readmissions in pediatric populations is relatively small, and efforts should continue to assure the highest-quality inpatient care and discharge planning," he said.
"Further research is needed to validate the applicability and appropriateness of PPR inclusions and exclusions in pediatric populations before it can be recommended for such consequential purposes as financial penalties to children’s hospitals."
During a discussion of the findings, an attendee asked whether this type of research was worth burning up so many bright minds, given that the data have repeatedly shown that pediatric readmission rates are so low. Dr. Gay responded, saying that, quite frankly, the work is necessary to avoid financial penalties and to drive home the point that quality improvement measures focused on pediatric readmissions are "bogus."
In a statement responding to Dr. Gay’s analysis, 3M Health Information Systems’ senior vice president of clinical and economic research, Richard Averill, said, "In payment applications, the number of PPRs for each provider is compared on a risk-adjusted basis to identify those providers with an ‘excess’ number of readmissions that are potentially preventable. Any payment adjustment for readmission is based only on the extent of excess readmissions.
"If all providers had the same PPR rate, no provider would have any excess PPRs and there would be no payment adjustment for readmission. Unfortunately, Dr. Gay’s comments omit any mention of how PPRs are actually being used and that they are not intended to provide a direct measure of the number of preventable readmissions."
Mr. Averill noted that Dr. Gay’s analysis observes that "half of readmissions detected by an ‘all-cause’ method would not be considered potentially preventable using PPRs. An examination of excess PPR rates across children’s hospitals can be an important quality improvement tool," he said.
In a separate interview, Dr. Shawn Ralston, chief of the pediatrics section, Dartmouth-Hitchcock Medical Center, Lebanon, N.H., agreed with Dr. Gay and said there are more pressing issues in pediatric clinical quality that deserve attention.
"I want meaningful quality measures developed by people with pediatric expertise and not things extrapolated from adult patients," Dr. Ralston added.
Dr. Gay reported serving as the medical consultant to the Children’s Hospital Association.
AT PEDIATRIC HOSPITAL MEDICINE 2013
