Recent Studies
- A Pediatric Quality Measurement Center and Testing Laboratory
Funded by the Agency for Healthcare Research and Quality
PI: Jeffrey H. Silber, MD, PhD
Improving Process Measurement
Funded by the Agency for Healthcare Research and QualitysFunded by the Agency for Healthcare Research and Quality
PI: Jeffrey H. Silber, MD, PhDThis is a proposal from the Children's Hospital of Philadelphia (CHOP) to become part of the Pediatric QualityMeasures Program's (PQMP's) Centers of Excellence (CoE) network. CHOP has formed an organizational architecture for a new CoE that builds on a strong, existing infrastructure already in place including investigators with breadth and depth in leadership and methodological expertise in quality and outcome assessment, unique data resources, and strategic partnerships. The CoE will include five methodological cores: (1) Quality Informatics-using the tools of informatics to adapt quality metrics to the electronic health record (EHR); (2)Multivariate Matching Laboratory--The lab will provide a unique analytic strength that will augment traditional analytic approaches to create scientifically rigorous pediatric risk adjustment models, produce better assessment of health and healthcare disparities, and improve validation of new quality measures; (3) Hospital Metrics-development of measures that focus on inpatient quality; (4) Medical Home-measurement of the performance of the medical home and community services that connect to it; and, (5) Health Plans-a special emphasis on measurement of plan enrollment duration and stability. The proposal has 6 specific aims: (1)Form the organizational architecture for a new CHOP CoE devoted to advancing the science of quality measurement and participating effectively in the PQMP; (2) Enhance a core measure of pediatric healthcare quality-avoidance of inappropriate use of antimicrobial therapy for otitis media with effusion-by adapting it for electronic health records; (3) Develop and validate pediatric readmission outcome measures including a traditional readmission measure based on information obtained on admission and a new readmission measure based on information collected at discharge, with the intent of assessing outpatient care related to readmissions; (4) Develop and validate the duration and stability of child health insurance coverage measures; (5) Enhance the developmental screening core measure by adapting it for EHRs and linking screening data from primary care to receipt of community-based early intervention services; and, (6) Develop a Multivariate Matching Lab and apply matching methods to risk adjustment, disparity assessment, and metric validation. The CHOP CoE will collaborate with several children's hospitals in the Pediatric EHR DataSharing Network (PEDSNet) and the GE Healthcare Medical Quality Improvement Consortium multi-practice data warehouse of over 15 million electronic health records. Stakeholder alliances with State Medicaid Directors, Independence Blue Cross of Pennsylvania, clinicians, families, and others, will provide substantive input on the applicability and importance of all proposed metrics. In summary, with the expertise, data, infrastructure, and alliances that are necessary to be successful, the CHOP CoE proposal is highly responsive to the requirements of the U18 RFA-HS-11-001 and should provide unique value to the PQMP.
- Improving Process Measurement
Funded by the Agency for Healthcare Research and Quality
PI: Jeffrey H. Silber, MD, PhD
Improving Process Measurement
Funded by the Agency for Healthcare Research and QualitysFunded by the Agency for Healthcare Research and Quality
PI: Jeffrey H. Silber, MD, PhDThis proposal aims to improve process measurement in order to better assess hospital quality of care. Discrete process measures like those used by CMS for Hospital Compare (HC) are commonly used by healthcare evaluation organizations because they are: (1) quick to collect (require less time in abstraction than traditional global measures); (2) easy to understand and point to potentially improvable provider actions; and (3) purportedly require no severity adjustment. In fact, a common motivation for utilizing process measurement is the belief that such measures do not require severity adjustment, beyond coarse inclusion and exclusion criteria. This proposal examines two fundamental questions regarding process measurement: (1) Do discrete process measures such as those utilized in CMS Hospital Compare need severity adjustment above and beyond the selection criteria commonly used; and (2) Are there other sampling schemes that could be utilized rather than typical random sampling (as is done by CMS) that could better (i.e., more efficiently and with less bias) sample processes at hospitals and therefore allow for the collection of more global process measures that could have stronger associations with outcomes than the discrete measures utilized in Hospital Compare. We present a conceptual model to study the need for severity adjustment for process measures and to compare various process measurement schemes based on reducing overall mean square error (MSE): This study will propose and present preliminary data on a potentially better method to sample charts for process measurement based on a multivariate matching algorithm we call "Multivariate Template Matching" which produces directly standardized matches of patients in order to better select patients inside hospitals to compare process of care across hospitals. We will also examine process measures using a Medicare data set collected as part of the CMS Hospital Compare initiative to (a) establish that we can achieve excellent matches using the Template Matching algorithm; and (b) test whether the bias observed with the present sampling schemes used by CMS for Hospital Compare is reduced using Template Matching. In summary, working with a data set to which CMS has given us special permission to analyze, we will formally test for bias in Hospital Compare, and formally test a more optimal scheme for conducting process measurement through Multivariate Template Matching. If successful, Multivariate Template Matching would allow for the collection of more detailed global process measures since the required sample size for abstraction (and therefore cost) could be reduced.
- Impact of Obstetric Unit Closures on Pregnancy Outcomes
Funded by the Agency for Healthcare Research and Quality
PI: Scott A Lorch, MD, MSCE
Impact of Obstetric Unit Closures on Pregnancy Outcomes
Funded by the Agency for Healthcare Research and Quality
PI: Scott A Lorch, MD, MSCEWith over 4 million infants delivering in the United States every year, the provision of obstetric care is a critical part of the health care system of any locality. However, the reduction of obstetric services, especially through the closure of obstetric units, is a common occurrence. As with studies of hospital closures, the impact of obstetric unit closures on the health of the surrounding population is not well described. Our group has previously investigated the case of Philadelphia, which experienced a 40% reduction in obstetric beds between 1997 and 2005. Compared to a pre-closure period of 1995-1996, we found that these closures were associated with in an initial increase in neonatal mortality, a decline in the number of deliveries via Cesarean section, and a persistent increase in neonatal and maternal delivery complications. Although this single urban area experienced adverse effects from the closure of multiple obstetric units, it is not known whether similar results are possible in markets that experience the loss of fewer obstetric units; what the differential effects of an obstetric closure experienced by women in the same urban market; and how changes in the provision of obstetric care by the remaining obstetric units could modify these results. Given these questions, this proposal will (1) identify the short-term and long-term outcomes of obstetric unit closures on pregnancies; (2) determine women at greatest risk for adverse outcomes associated with obstetric unit closures; and (3) identify adaptive strategies for remaining obstetric units to ameliorate these changes in pregnancy outcomes. To answer these questions, this study will use hospital discharge records linked to birth and death certificates for over 10 million deliveries occurring in three states between 1995 and 2007 to construct two analytic models. First, this study will construct a hospital choice model to identify the patients or geographic areas at risk for a direct effect of an obstetric unit closure. Second, the study will use a difference-in-differences approach with multiple time frames to control for secular trends in each outcome and control for case mix and stable outcome differences between the geographical areas undergoing a reduction in obstetric supply and those that are not. This research falls under the Value portfolio of the Agency for Healthcare Research and Quality research program, as it projects how changes to the organization of the obstetric market may affect the quality and value of this system. Improved understanding of the impact of such service reductions in the obstetric market will result in more efficient use of health care services and optimize the value of health care dollars spent on obstetric care.
- Understanding Racial Disparities in Surgical Outcomes
Funded by the National Institute of Aging
PI: Jeffrey H. Silber, MD, PhD
Understanding Racial Disparities in Surgical Outcomes
Funded by the National Institute of Aging
PI: Jeffrey H. Silber, MD, PhDMany studies have documented important racial disparities in surgical outcomes but their etiologies are not clear. While we generally find better surgical outcomes (lower mortality) at teaching hospitals, and a higher rate of minority patients at these hospitals, the survival benefit at teaching hospitals does not seem to equally apply to white and black patients. This study will explore why we observe disparities in three common surgical specialties that are performed throughout most hospitals where surgery is practiced: general, orthopedic, and vascular surgery.
The project will focus on a classic measure of surgical practice - the procedure time. We will use Medicare claims to estimate procedure length in a population of Pennsylvania patients. While always available on chart review, the ability to study surgical procedure time on a vast scale with Medicare claims is new. Anesthesiologists began billing Medicare by the minute in 1994. Using these bills to estimate procedure length in a population of Pennsylvania patients, we have recently published that there was a significant and clinically important disparity in the length of operations based on race and income, after detailed adjustment for patient characteristics and procedure type. This disparity was greatest at hospitals with residency programs, where the procedure time of black patients in the lowest third of the income distribution was on average 31 minutes longer than that of white patients at the same level of income (P<0.0001). In non-resident hospitals the mean gap was 8 minutes (P<0.0001). In further work we have shown that the disparity in procedure length gets larger with longer procedure times (at the 95%ile using quantile regression, the difference between blacks and whites is about ¾ of an hour). Furthermore, in our examination of all teaching and non-teaching hospitals in the state of Pennsylvania, the 15 hospitals with the largest number of black patients undergoing orthopedic and general surgery, excess adjusted procedure time (black-white by hospital) was statistically significant in 5 of 15 hospitals, with their mean excess times ranging from 9 to 16 minutes. The strength of using operative time as a measure of disparity is that it reflects precisely what the surgical and anesthesia team have done, and is not confounded by patient compliance or preferences, as some other measures of process and outcome may be.
This project aims (1) To study and describe racial and income/SES disparity in procedure time across surgeons and hospitals; (2) To identify the determinants of disparity in procedure time across surgeons and hospitals. We will study those geographic, hospital, and surgeon characteristics associated with the greatest disparities in procedure time; and (3) To make practical recommendations to departments of surgery regarding procedure length and to create models to allow surgery departments to practically monitor such disparities in order to develop mechanisms to eliminate problems if found.
- Describing and Understanding Racial Disparities
Funded by the Agency for Healthcare Research and Quality
PI: Jeffrey H. Silber, MD, PhD
Describing and Understanding Racial Disparities
Funded by the Agency for Healthcare Research and Quality
PI: Jeffrey H. Silber, MD, PhDThis study explores a new method for describing and understanding racial disparities through the use of "Tapered Multivariate Matching" (TMM). TMM represents a new theory and conceptual framework for examining racial disparities that will aid in better measuring and more clearly describing racial disparities as well as better isolating and identifying specific features of the healthcare system that lead to differences in quality of care across racial groups. Presently, the size of a disparity is typically measured by reporting a regression model coefficient on race. However: Minorities are, by definition, underrepresented in these models because whites generally outweigh blacks in any analysis, which can tend to weight model coefficients from a white perspective. At the same time, minorities often present with very different covariate distributions; therefore, the interactions needed to properly define the relationships between race and explanatory covariates are usually not analyzed, or if analyzed, not reported. It therefore becomes very difficult for policy analysts to clearly convey to the public the true extent of the racial disparity. TMM starts with the minority population (say the black or Hispanic population) and, using multivariate matching, matches white patients to the entire minority group. Matching is done sequentially: for example, starting matching at diagnosis, then matching at both diagnosis and treatment. In so doing, TMM can, for example, directly compare survival for whites who present at diagnosis similarly to their matched black pairs (using again all blacks), and then compare whites who were matched on both diagnosis and treatment variables, again matched to the same black patients (all blacks). For reporting, TMM can decompose Total Racial Disparity at, say, 5 years from diagnosis (DTOT) into the sum of the disparity in survival due to diagnosis differences between whites and blacks (DDx) plus differences in survival due to initial treatment (DTx) plus a residual disparity (DR), or DTOT = DDx + DTx + DR. These components are more clearly understood than reporting a complex, fully interacted model, and they provide direct information on aspects of the treatment of patients that can be changed to improve quality.
This study will apply the TMM framework to breast and colorectal cancer (though TMM can be used on any medical problem beyond cancer) using the national Medicare-SEER database and demonstrate the usefulness of TMM in both identifying specific aspects of patient care that lead to quality differences by race, and to demonstrate how TMM results can be conveyed in a clear and understandable report that can be used to help reduce racial disparities. Furthermore, the project will create a website to facilitate the understanding and implementation of TMM. Through TMM, we aim to establish a new and better method to measure and describe disparities in patient treatment and outcomes for all areas of medicine and in so doing, shed light on etiologies for these quality differences and on how to reduce them.
- Scope of Services for Fetal Mortality: Epidemiology and Performance Measurement
Funded by the Center for Disease Control
PI: Scott A. Lorch, MD, MSCE
Scope of Services for Fetal Mortality: Epidemiology and Performance Measurement
Funded by the Center for Disease Control
PI: Scott A. Lorch, MD, MSCENeonatal mortality rates have been proposed as a measure of the quality of neonatal care. There has been little validation of the correct metric for their use. In Great Britain, hospitals receive raw mortality rates, unadjusted for hospital casemix. Studies from the Vermont Oxford Network use neonatal mortality rates alone, while studies from Ciaran Phibbs include both neonatal mortality rates and "preventable" fetal deaths. These "preventable" fetal deaths are obtained from ICD-9 codes contained on maternal hospital records linked to additional codes contained on the fetal death record.
There are several advantages and disadvantages to the inclusion of fetal deaths in any mortality metric. Among the advantages include poor quality hospitals may have poor resuscitation practices, converting infants who would be neonatal deaths at higher quality hospitals to fetal deaths at lower quality hospitals. At the least, this conversion would decrease the perceived quality differences between high and low quality hospitals, at the worst this conversion could reverse the assessed quality at these types of hospitals; including only "preventable" fetal deaths eliminates the fetal deaths that are inevitable, either because the fetus was dead upon admission to the hospital or because of pre-existing condition. However, fetal deaths are poorly collected, especially reasons for the demise; there are no good validated measures of "preventable" fetal deaths.
There are additional aspects of the epidemiology about fetal deaths that are important to understand from a public health perspective. First, there are limited studies that define risk factors for fetal deaths beyond specific medical conditions experienced by the mother during pregnancy. Second, there is little information on geographic variations in fetal deaths, across different types of urbanization or the location of treatment. One concern is that "non-preventable" fetal deaths may be distributed unequally across hospitals, leading to biased assessments of the care provided by these hospitals. Finally, it is not known how fetal deaths explain racial disparities in all-mortality infant rates.
The purpose of this study is (1) to understand the epidemiology of fetal deaths, as identified through vital statistics data linked to hospital discharge records, and (2) to determine the importance of fetal deaths when using mortality as a measure of the performance of neonatal or perinatal care. To do this, we will calculate 4 separate mortality statistics for each hospital included in our linked birth/death certificate/maternal and infant hospital discharge record database, which includes information from over 8,000,000 births in California, Missouri, and Pennsylvania between 1995 and 2005.
- Parental Trust and Racial Disparities in the Care of Discharged Premature Infants
National Institute of Child Health and Human Development
PI: Scott A. Lorch, MD, MSCE
Parental Trust and Racial Disparities in the Care of Discharged Premature Infants
Funded by National Institute of Child Health and Human Development
PI: Scott A. Lorch, MD, MSCEThe care of premature infants is a rapidly growing public health concern in the United States, with over 57,000 infants born every year with a birth weight under 1500 grams. With recent advances in neonatal care, more premature infants are surviving to discharge from the neonatal intensive care unit (NICU). Even though this high-risk group of patients is particularly sensitive to the care they receive after discharge, relatively little is known about the quality of such post-discharge care, including the receipt of preventive services, frequency of emergency room visits, and the adequacy of immunizations. For those few outcomes that have been studied, data suggest that there are racial disparities in the care received by premature infants after discharge. Although racial disparities in health care have been widely described for over twenty years, the root causes for such racial disparities in health care are only beginning to be understood. While some disparities appear to result from attendance at lower quality inpatient and outpatient facilities, disparities often persist within a specific health care site. The Institute of Medicine report on Unequal Treatment highlights multiple factors that may contribute to disparities within a health care site. Many of these factors are aspects of the doctor-parent relationship such as communication styles, expectations, and trust.
Over the last several years, our research team has focused on the contribution of health care related trust and distrust to racial disparities in health care. Patient trust in the health care system and health care providers represents an important focus for understanding racial disparities in health care - both because of the evidence supporting its role in these disparities and because it may provide a feasible lever of change.
Given the current lack of insight into racial disparities in the quality of care delivered to premature infants after discharge from the neonatal intensive care unit, this project proposes (1) to identify the relative contribution of race to variations in process and outcome measures of post-discharge care for the premature infant, and (2) to examine the contribution of health care-related distrust to racial disparities in post-discharge care, placing health care distrust within the broader context of other patient, provider and system factors that may contribute to such disparities. The overall hypothesis of the project is that after controlling for other factors, the level of parental distrust in the medical system contributes to racial disparities in the quality of care received by discharged premature infants. The project will examine a large cohort of discharged premature infants across 32 outpatient pediatric sites affiliated with The Children's Hospital of Philadelphia (CHOP).
- The Organization of Emergency Medical Services for Improved Public Health
Leonard David Institute of Health Economics
Co-PIs: Tanguy Brachet, PhD; Guy David, PhD
The Organization of Emergency Medical Services for Improved Public Health
Funded by Leonard David Institute of Health Economics
Co-PIs: Tanguy Brachet, PhD; Guy David, PhDThis study spans a number of cross?school and cross?disciplinary health services research projects intended to enhance our understanding of the role, structure, efficiency, and efficacy of Emergency Medical Services (EMS). More specifically, we are interested in four broad topics tied together by a rich and unique dataset detailing over four million prehospital medical and trauma interventions in the state of Mississippi over a period spanning a decade and a half. The four areas we aim to study are the profile and benefits of human capital accumulation among emergency medical technicians and paramedics; the evaluation of prehospital clinical practice and its effect on patient outcomes; the determinants and consequences of EMS privatization; and the determinants of child safety seat use, and its impact on injuries, EMS resources and intensity.
The U.S. public health system relies on EMS networks to limit the casualties from such incidents as myocardial infarctions, motor vehicle crashes, and acts of violence. They also provide access to emergency care to the uninsured, and so form part of the safety net. More than three decades after Congress passed the Emergency Medical Services Systems Act of 1973, which delegated responsibility to municipalities and established the first standards for prehospital emergency care, patients are treated by EMS with little or no evidence that the care they provide is optimal. In EMS settings across the country, there is significant variation in legislation, contracts, system financing, organization of EMS providers and payers, certification of EMS personnel, dispatch, and performance measurements. To date, there exists no systematic research on the delivery of EMS, the determinants of their organizational structure, or the effect of human capital accumulation on performance. The U.S. EMS system is complex and varied in structure, with sparse national standards and numerous models of delivery. The largest EMS providers in the nation (i.e. AMR, Rural/Metro, Emergystat, etc.) have a substantial and growing presence in Mississippi, which contributes to the external validity of these studies, which aim to fill some of the holes in the understanding of EMS organization and effectiveness by exploiting a unique dataset obtained from the Office of Emergency Planning and Response at the Mississippi Department of Health. - Perinatal Regionalization and Quality of Care
Funded by the Agency for Healthcare Research and Quality
PI: Scott A. Lorch, MD, MSCE
Perinatal Regionalization and Quality of Care
Funded by the Agency for Healthcare Research and Quality
PI: Scott A. Lorch, MD, MSCEVery-low-birth weight (VLBW) infants with a birth weight less than 1,500 grams are a significant public health concern, comprising 57,000 deliveries per year at a cost of over 4 billion dollars for medical care. Regionalization policies have been developed to optimize the care of these high-risk newborns, centralizing the care of VLBW infants at a few specialized hospitals within a defined geographic region. However, these policies have recently weakened in many areas of the U.S.
How these shifts in policy have affected the quality of neonatal care and patient safety is unknown, since prior work has been limited by selection bias; poor control of casemix severity; and the lack of a uniform measure of the degree of regionalization in the areas of interest. To address these methodological issues, this study will use an instrumental variables approach to answer the following specific aims: (1) to quantify the difference in the quality of neonatal care between hospitals; (2) to determine the effect of perinatal de-regionalization on the rates of mortality, failure-to-rescue, complication and 21-day readmission in VLBW infants; and (3) to develop a predictive model that explains neonatal outcomes as a function of the quality of neonatal care and regionalization policy in a given geographical area.
This project will use population data from the years 1992 to 2002 in three states that have an estimated 12,800 VLBW births per year. Two-stage regression techniques will be used to control for measured confounders, while various instruments including the differential travel time from the residential zip code to regional center and to the nearest delivery hospital will be used to control for unmeasured bias. Results from this study will allow policy makers to develop perinatal regionalization policies for individual areas after accounting for geographic differences in hospital quality and population, improving the quality and safety of care for VLBW infants. This improvement in the quality of neonatal care delivered to this high-risk population of infants should result in short and long-term improvements in the public health, outcome, and medical-related costs of caring for VLBW infants. - Obesity and Surgical Outcomes
Funded by the National Institute of Diabetes and Digestive and Kidney Diseases
PI: Jeffrey H. Silber, MD, PhD
Obesity and Surgical Outcomes
Funded by the National Institute of Diabetes and Digestive and Kidney Diseases
PI: Jeffrey H. Silber, MD, PhDThe overall aim of the Obesity and Surgical Outcomes study (OBSOS) is to improve the outcomes of obese patients who must undergo general, orthopedic, vascular or urological surgery. Obesity has been implicated in predisposing patients to worse surgical outcomes (death and complications). However, the etiology of these outcome differences is not well understood or well defined. Developing approaches to improving outcomes in this population is therefore challenging.
In this study we will (1) determine how obesity influences the post-operative outcomes of death, complications, failure-to-rescue (the probability of death after complications), length of stay, readmission rates and cost. We will also (2) identify potential risk factors that may be especially important in treating obese as compared to non-obese patients, such as control of serum glucose, use of preoperative antibiotics, early ambulation and epidural analgesia. In addition, we plan to (3) determine whether there are disparities in the care of obese patients undergoing surgery as compared to normal weight patients, such as differences in the surgical procedures offered, anesthesia technique and postoperative pain control.
The study will use an efficient design that incorporates a three state, 50 hospital, Medicare augmented claims analysis that includes height, weight, and serum albumin information from charts (N=20,000), and a nested, detailed chart abstraction (N=4,000) using a matched cohort design to answer questions that require clinical data not available from Medicare claims. The multi-institutional OBSOS study will examine differences in outcomes and treatments between obese and normal weight patients, and in so doing, will make important contributions to the care of these patients. - Infant Functional Status and Discharge Management
Funded by the Maternal and Child Health Bureau
PI: Jeffrey H. Silber, MD, PhD
Infant Functional Status and Discharge Management
Funded by the Maternal and Child Health Bureau
PI: Jeffrey H. Silber, MD, PhDDischarge algorithms for premature infants are controversial because (1) they determine the allocation of significant resources; (2) discharge decisions often place families and physicians at odds with insurers; and (3) there is a lack of information on the relationship between infant physiologic status at discharge and subsequent resource utilization and outcomes. This project seeks to develop a more optimal algorithm to support infant discharge decisions, based on the relationship between physiologic parameters reflecting infant functional status and total resource utilization associated with these discharge decisions. The main hypothesis to be tested is whether there exists a range of discharge algorithms in which longer hospital stays may result in reduced overall resource utilization.
Using a unique database maintained by the Kaiser Permanente Medical Care Program (KPMCP) among six hospitals in Northern California, the study will (1) randomly select a cohort of 1,400 premature infants; and (2) abstract charts to extract key daily physiologic data on feeding, weight gain, temperature stability, respiratory status and medical stability. Based on these data, investigators will (3) perform a cost accounting starting from a consistent point in each hospitalization (defined as the point when a minimal level of maturity is achieved and discharge is first possible, and extending to six months after achieving such physiologic maturity); and (4) construct models to predict total cost and readmission rate as a function of specific patterns of discharge defined by specifying physiologic parameters associated with functional status. Finally, investigators will (5) determine optimal patterns of discharge using the data collected and the derived economic and outcomes model.
Once models are developed and the cost accounting is complete, the study will formally test whether a specific discharge algorithm will be optimal in terms of minimizing cost and readmissions. Results from the study will ultimately provide important information relating physiologic status to outcomes and cost that can be use to develop more rational discharge algorithms. - Aggregated Complication Measure for Neonatal Quality of Care
Funded by the Maternal and Child Health Bureau
PI: Scott A. Lorch, MD, MCSE
Aggregated Complication Measure for Neonatal Quality of Care
Funded by the Maternal and Child Health Bureau
PI: Scott A. Lorch, MD, MCSEEvery year in the U.S. more than 57,000 infants are delivered at least two months prematurely. Many of these infants are severely disabled, and their estimated cost for medical care to age 18 is between $4 and 5.4 billion. A large part of these long-term costs result from complications developed after birth in the neonatal intensive care unit (NICU).
While some outcomes are an unavoidable consequence of premature birth, increased cost and long-term disabilities may result from the quality of care these infants received in the NICU. However, there are no validated measures of the quality of neonatal care. Currently available measures, such as mortality and individual complication rates, lack adequate power and validity and are confounded by influences outside the control of the NICU.
This study will develop and validate a new quality measure, the aggregate complication measure (ACM), to evaluate hospitals according to the quality of care they provide after using direct standardization methods to control for differences in casemix. This study will use population data from Pennsylvania, New York, Missouri and California, areas that have a combined estimated 95,000 premature births each year. Statistical modeling techniques will determine valid weights for each complication based on the impact of each complication on one of four outcomes of interest: death, length of stay, cost or 28-day readmissions. Completion of this project will produce a valid measure of neonatal quality to both evaluate hospitals and quantify the difference in quality between hospitals. Improved measures of neonatal quality will help facilitate initiatives to lower costs and improve the quality of neonatal care by reducing unnecessary morbidity and mortality in these infants. - Impact of Resident Work Hour Rules on Errors and Quality
Funded by the National Heart, Lung, and Blood Institute
PI: Kevin G. Volpp, MD, PhD, Co-PI: Jeffrey H. Silber, MD, PhD
Impact of Resident Work Hour Rules on Errors and Quality
Funded by the National Heart, Lung, and Blood Institute
PI: Kevin G. Volpp, MD, PhD, Co-PI: Jeffrey H. Silber, MD, PhDThe Accreditation Council for Graduate Medical Education (ACGME) released rules effective July 1, 2003 that affected duty hours for all ACGME-accredited residency programs in all specialties. These rules represent the largest national effort to reduce medical errors since the publication of the Institute of Medicine's "To Err is Human" in 2000, and will directly impact the healthcare received by the 44 percent of patients cared for in teaching hospitals in the U.S.
This project will evaluate the effect of the duty hour rules on patient safety and quality of care, utilizing national data available though Medicare and pre-validated measures of quality including the Agency for Healthcare Research and Quality (AHRQ) Quality Indicators. The primary aims of the project are to compare changes in the rate of mortality and failure-to-rescue (death after complications) in teaching hospitals and non-teaching hospitals before and after implementation of the ACGME work hour rules. The secondary aims are to examine differences in AHRQ Patient Safety Indicators (PSIs) before and after the rule change and to study how length of stay (LOS), the probability of a prolonged length of stay and conditional length of stay (LOS once a stay is prolonged) changed in teaching vs. non-teaching hospitals.
The study will be based on approximately 16 million surgical and medical admissions collected from the Medicare's MEDPAR data set spanning the years 1999 through 2003 (before the rule change) and 2003 through 2005 (after the rule change). We will use a multiple time series design with non-teaching hospitals as a control for teaching hospitals and will examine how effects differ in accordance with hospital dependence on residents (resident/bed ratio), program type and size and hospitals' baseline financial status. Results from this study will be central to any future efforts to reduce errors in teaching hospitals through resident work hour reform. - Impact of Resident Work Hours on Errors and Quality in VA Hospitals
Funded by the Department of Veterans Affairs
PI: Kevin G. Volpp, MD, PhD, Co-PI: Jeffrey H. Silber, MD, PhD
Impact of Resident Work Hours on Errors and Quality in VA Hospitals
Funded by the Department of Veterans Affairs
PI: Kevin G. Volpp, MD, PhD, Co-PI: Jeffrey H. Silber, MD, PhDApproximately 70 percent of Veterans Affairs (VA) hospitals are teaching hospitals, and much of the direct care provided to patients is the responsibility of medical and surgical residents. In recent years, there has been an emerging consensus that acute and chronic sleep deprivation may contribute significantly to medical errors among residents. In response to the outcry over medical errors, the Accreditation Council for Graduate Medical Education (ACGME) released rules effective July 1, 2003 that affect duty hours for all ACGME-accredited residency programs in all specialties.
The effects of these duty hour limitations on medical errors and other measures of quality of care are important to understand because they affect the care of the vast majority of inpatients treated in VA hospitals each year. The overall goal of this proposal is to improve patient safety and quality of care by creating an evidence base of the positive and negative effects of the ACGME duty hour reforms. This will inform further efforts to reduce errors in VA teaching hospitals through resident work hour reform. - Provider Specialty and Outcomes in Ovarian Cancer
Funded by the National Cancer Institute
PI: Jeffrey H. Silber, MD, PhD
Provider Specialty and Outcomes in Ovarian Cancer
Funded by the National Cancer Institute
PI: Jeffrey H. Silber, MD, PhDThe goal of this project was to study practice variations and outcomes across specialty type in ovarian cancer, and to develop an approach for conducting chart review in the SEER-Medicare data base that takes into consideration issues of patient confidentiality. In particular, the project studied how the specialty of a healthcare provider influences surgery and chemotherapy, and how such differences influence mortality in ovarian cancer. The project used multivariate counter-matching and conditional logistic regression, in order to control for chemotherapy and surgical treatment in a manner not previously accomplished.
Counter-matching allowed for adjustment by factors such as age, patient co-morbidities, stage, year, chemotherapy (as determined from Medicare claims) and other provider information when appropriate, while utilizing the minimum number of pairs needed to detect important differences. Matching was complemented with regression modeling using all available claims data in order to utilize maximum information available. The aims of the study were to determine if outcomes after chemotherapy differ with the specialty of the provider delivering the chemotherapy; chemotherapy intensity differs across the specialty of the provider delivering that chemotherapy; survival is a function of the type of surgical specialist performing the primary cancer surgery. Also, the project aimed to study the referral patterns associated with different surgical specialties; and develop a Phase-II minimally intrusive approach that is sensitive to the concerns of forthcoming federal regulations regarding patient privacy, while allowing for important health services research questions to be answered using the data resources of the SEER-Medicare data base with chart review. Our hope is that this study will serve as a model for other matched cohort studies involving chart review utilizing the SEER-Medicare database. - Race, Treatment and Endometrial Cancer Survival
Funded by the National Cancer Institute
PI: Katrina Armstrong, MD, MSCE
Race, Treatment and Endometrial Cancer Survival
Funded by the National Cancer Institute
PI: Katrina Armstrong, MD, MSCESurvival after the diagnosis of endometrial cancer varies significantly between African-American and Caucasian women. Between 1992 and 1998, five-year survival for African-American women with endometrial cancer was 58.9 percent, compared to 85.8 percent for Caucasian women. Prior studies have identified several factors that contribute to this disparity, including later stage at diagnosis and higher tumor grade. However, significant differences in survival between African-American women persist even after adjusting for these factors.
There are several reasons to believe that differences in the prevalence and characteristics of treatment may contribute to this residual survival disparity. African-Americans are less likely to undergo definitive treatment for many different surgical and medical conditions. Characteristics of treatment (including provider characteristics, hospital characteristics and aggressiveness of therapy) are associated with outcome in many different settings, including surgery for lung, pancreatic and breast cancer.
Understanding the prevalence and outcomes of differences in treatment characteristics between African-American and Caucasian women with endometrial cancer offers a potentially promising new approach to understanding and addressing the disparity in survival. In this study we will use SEER Medicare linked data to examine the outcomes of African-American and Caucasian women diagnosed with endometrial cancer between 1991 and 2000. The primary outcomes will be three-year survival rates and times (which may be censored) as tracked by Medicare vital statistics. Analyses will adjust for comorbidity, socioeconomic status and tumor characteristics.
Our three specific aims explore the contribution of differences in treatment to the higher mortality among African-American women. We categorize differences in treatment into three categories: differences in the rates of treatment, differences in the extent or aggressiveness of treatment, and differences in the characteristics of the providers or hospitals (including surgical specialty, volume and teaching status). For each category, we will explore the differences between African-American and Caucasian women, the association with outcome and to what extent the factors explain the excess mortality among African-American women. - Development of a Risk Adjusted Aggregated Complication Measure
Funded by the Pennsylvania Health Care Cost Containment Council
PI: Jeffrey H. Silber, MD, PhD
Development of a Risk Adjusted Aggregated Complication Measure
Funded by the Pennsylvania Health Care Cost Containment Council
PI: Jeffrey H. Silber, MD, PhDThis project will develop a method to aggregate across complications in order to produce more stable measures. The study will examine mortality, average lengths of hospitalization, length of stay outlier rates and ratings, readmission rates for any reason and for complication/infection and regionally adjusted average hospital charges. These measures will be more stable both because they will aggregate larger numbers of complications per diagnosis-related group and per hospital, and also because the aggregated complications can be severity-adjusted and used directly to rank and evaluate Pennsylvania hospitals. The resulting weighted and aggregated complication measure can then be used to rank hospitals. These rankings can be tested for construct validity by comparing them to mortality rates and to standard complication rates.
- Investigating the Relationship Between Measures of Hospital Occupancy and Length of Stay
Funded by the Leonard Davis Institute of Health Economics
PI: Scott A. Lorch, MD, MSCE
Investigating the Relationship Between Measures of Hospital Occupancy and Length of Stay
Funded by the Leonard Davis Institute of Health Economics
PI: Scott A. Lorch, MD, MSCEThe impact of hospital occupancy on the outcomes of patients is important for the organization of hospital delivery systems and the quality of inpatient care. This issue has been difficult to study because of several unique properties of hospital occupancy and length of stay. Important variables, such as (1) the decision to admit and discharge a patient and (2) daily nurse staffing decisions made in response to overcrowding, are omitted from most studies because standard population-based administrative data do not include enough data to construct these measures. Also, hospital occupancy and length of stay could be determined simultaneously and bias results obtained by conventional regression techniques.
However, the neonatal population offers a potentially unique approach to the study of this issue. Using detailed data from Kaiser Permanente hospitals, this project will obtain unbiased measures of the relationship between occupancy and length of stay by (1) including information on daily staffing and severity of illness at discharge not available in most administrative datasets and (2) using various methods to account for simultaneity including time-varying models, instrumental variables, and proxy measures for length of stay.
By determining valid measures of hospital occupancy and understanding how hospital overcrowding can affect length of stay, this project will serve as critical pilot data for future projects investigating the ways that hospitals and units adapt to hospital overcrowding and whether overcrowding results in early discharge of neonatal patients. - The Impact of Physiological Maturation of the Premature Infant on Perinatal Outcomes
Funded by the University of Pennsylvania McCabe Foundation
PI: Scott A. Lorch, MD, MSCE
The Impact of Physiological Maturation of the Premature Infant on Perinatal Outcomes
Funded by the University of Pennsylvania McCabe Foundation
PI: Scott A. Lorch, MD, MSCEThe goals of this project are to (1) determine the rate of physiological maturation of premature infants in those skills necessary for discharge to home (degree of respiratory support, control of breathing, coordination of feeding, ability to gain weight and control of core body temperature); (2) identify the relationship between the attainment of these skills and the short- and long-term outcome measures of 1-, 3- and 12-month readmission, total cost of medical care and use of emergency room services; and (3) validate a function maturity score to assist clinicians in optimizing the timing of discharge of premature infants.
We hypothesize that (1) attainment of temperature control and breathing will occur closer to discharge for infants
28 weeks gestation, while attainment of feeding and weight gain will occur closer to discharge for infants between 29 and 34 weeks gestation; (2) the rate of attainment of feeding and control of breathing will be associated with a decreased rate of readmission and cost of care; and (3) attainment of these skills at time of discharge will be association with a greater decrease in readmission, costs and emergency room use for infants 28 weeks gestation compared to infants between 29 and 34 weeks gestation. - An Economic Study of Adverse Events Across Conditions
Funded by Amgen
PI: Jeffrey H. Silber, MD, PhD
An Economic Study of Adverse Events across Conditions
Funded by Amgen
PI: Jeffrey H. Silber, MD, PhDAdverse events associated with chemotherapy represent a large, but not well quantified, risk of cancer treatment. While specific studies have investigated the costs of these adverse events for isolated diseases, no existing models take a wider, global perspective, looking at the risks and costs of adverse events across diseases and chemotherapeutic regimens.
The aim of this project was to develop a global model to estimate adverse events accounting for the distribution of cancer types. In particular, we studied the relationship between adverse effects such as neutropenia and anemia, chemotherapy usage and cancer types though a process of direct standardization. This work allowed for future comparisons across institutions on the management of adverse events given chemotherapy regimen and cancer types. - Improving Pediatric Severity Adjustment for Measuring Quality of Care
Funded by The Children's Hospital of Philadelphia
PI: Jeffrey H. Silber, MD, PhD
Improving Pediatric Severity Adjustment for Measuring Quality of Care
Funded by The Children's Hospital of Philadelphia
PI: Jeffrey H. Silber, MD, PhDThe aim of this project is to utilize the internal data being generated by The Children's Hospital of Philadelphia and its satellite practices to develop a minimum, uniform pediatric research database. This Children's Hospital initiative will develop the standards for each element to be included in the pediatric health services database. If adopted by other institutions, this minimum pediatric database will enable pediatric health services researchers to conduct multi-institution quality studies. Specifically, this project will (1) modify data collection systems to include metric ascertainment required by those developing quality of care models; (2) develop methodology to assess trends in pediatric quality of care; and (3) facilitate the sharing of data to improve benchmarking efforts with appropriate severity adjustment.
- Copyright © 2013, The Children's Hospital of Philadelphia.
- All rights reserved.
- Privacy Policies
- Terms of Use
- Children's Hospital of Philadelphia
- CHOP Research Institute