Medical Diagnosis-related Groups Research Articles

Comparing Hospital Length of Stay Risk-Adjustment Models in US Value-Based Physician Payments.

Under the Affordable Care Act, the US Centers for Medicare & Medicaid Services created the Physician Value-Based Payment Modifier Program and its successor, the Merit-Based Incentive Payment System, to tie physician payments to quality and cost. The addition of hospital length of stay (LOS) to these value-based physician payment models reflects its increasing importance as a metric of health care cost and efficiency and its association with adverse health outcomes. This study compared the Centers for Medicare & Medicaid Services-endorsed LOS risk-adjustment methodology with a novel methodology that accounts for pre-hospitalization clinical, socioeconomic status (SES), and admission-related factors as influential factors of hospital LOS. Using the 2014 New York, Florida, and New Jersey State Inpatient Database, we compared the observed-to-expected LOS of 2373102 adult admissions for 742 medical and surgical diagnosis-related groups (DRGs) by 3 models: ( a ) current risk-adjustment model (CRM), which adjusted for age, sex, number of chronic conditions, Elixhauser comorbidity score, and DRG severity weight, ( b ) CRM but modeling LOS using a generalized linear model (C-GLM), and (c) novel risk-adjustment model (NRM), which added to the C-GLM covariates for race/ethnicity, SES, discharge destination, weekend admission, and individual intercepts for DRGs instead of severity weights. The NRM disadvantaged physicians for fewer medical and surgical DRGs, compared with both the C-GLM and CRM models (medical DRGs: 0.49% vs 13.17% and 10.89%, respectively; surgical DRGs: 0.30% vs 13.17% and 10.98%, respectively). In subgroup analysis, the NRM reduced the proportion of physician-penalizing DRGs across all racial/ethnic and socioeconomic groups, with the highest reduction among Whites, followed by low SES patients, and the lowest reduction among Hispanic patients. After accounting for pre-hospitalization socioeconomic and clinical factors, the adjusted LOS using the NRM was lower than estimates from the current Centers for Medicare & Medicaid Services-endorsed model. The current model may disadvantage physicians serving communities with higher socioeconomic risks.

Value-based care through postacute home health under CMS PACT regulations.

To assess in a Medicare Advantage population (1) whether discharge to home health, compared with discharge to home, following an inpatient stay subject to CMS postacute care transfer (PACT) regulations, is associated with better outcomes or lower expenditures and (2) whether the impact differs among subpopulations. Claims-based retrospective cohort study. Instrumental variable (IV) analysis, with prior hospital-level probability of discharge to home health as the IV, to control for unobservable as well as observable confounders. Compared with 15,071 patients discharged to home, 4160 patients discharged to and receiving timely home health services were 60% less likely to be readmitted within 30 days and 37% less likely at 90 days. Total expenditures from time of admission to 90 days post discharge were 11% lower in the home health group. The association of discharge to home health with reduced readmission and reduced costs varied by subpopulations defined by surgical vs medical diagnosis-related group and receipt of intensive care management following discharge. The PACT policy may be promoting greater value by reducing readmissions while lowering total expenditures for patients who do not require intensive postacute care. Findings were in contrast to those of previous studies, in which discharge to home health has been associated with higher rates of readmission. Earlier studies did not control for unmeasurable confounders, involved narrowly defined populations, and used older data.

Heart failure hospitalizations and costs in ICD/CRT-D recipients following replacement or upgrade: the DECODE registry.

AimsThe aim of this study is to report heart failure hospitalization (HFH) rates and associated costs within 12 months following implantable cardioverter defibrillator (ICD)/cardiac resynchronization therapy defibrillator (CRT‐D) device replacement or upgrade from ICD to CRT‐D.Methods and resultsThe DEtect long‐term COmplications after icD rEplacement (DECODE) was a prospective, single‐arm, multicentre cohort study that explored complications in ICD/CRT‐D recipients. All clinical and survival data at 12 months were prospectively analysed. For each adjudicated HFH, admission and discharge dates and ICD‐9‐CM diagnosis and procedure codes were recorded. The reimbursement for each HFH was calculated for each diagnosis‐related group code. Between 2013 and 2015, 983 patients (mean age 71 years, male 76%, mean left ventricular ejection fraction 35%, and New York Heart Association Class I/II 75.6%) were enrolled. Patients underwent device replacement (900; 91.6%, 446 ICD/454 CRT‐D) or ICD upgrade to CRT‐D (83; 8.4%). Post‐replacement hospitalizations occurred in 220 patients, with the primary discharge diagnosis identifying cardiovascular causes in 175 patients (80%). Fifty‐five (5.6%) patients experienced at least one HFH. Overall, 91 HFH events occurred (9.6% event rate, 95% confidence interval: 7.7–11.7) in 70 patients; 66 (6.7%) patients died, 40 (60.6%) of cardiovascular causes. The HFH rate was significantly higher following upgrades, and the occurrence of HFH was associated with an 11‐fold increased mortality risk (95% confidence interval: 5.9–20.5, P < 0.0001). Medical diagnosis‐related group accounted for 91.2% of HFH; the mean cost per HFH was €5662 ± 9497, and the mean cost per patient was €9369 ± 12 687. On multivariate analysis, predictors of HFH were atrial fibrillation, chronic kidney disease, and all‐cause hospitalization within 30 days prior to the procedure.ConclusionsIn the DECODE registry, HFH and mortality rates in the year following ICD/CRT‐D replacement or upgrade were low. In this particular subset, underlying cardiac disease was the main driver of HFH, mortality, and higher healthcare expenditures.

Bending the cost curve: time series analysis of a value transformation programme at an academic medical centre

BackgroundReducing costs while increasing or maintaining quality is crucial to delivering high value care.ObjectiveTo assess the impact of a hospital value-based management programme on cost and quality.DesignTime series analysis of...

Bundled Payments for Care Improvement: Preparing for the Medical Diagnosis-Related Groups.

BACKGROUNDThe Centers for Medicare and Medicaid Services Innovation Center introduced the Bundled Payments for Care Improvement (BPCI) initiative in 2011 as 1 strategy to encourage healthcare organizations and clinicians to improve healthcare delivery for patients, both when they are in the hospital and after they are discharged. Mercy Health Saint Mary’s, a large urban academic medical center, engaged in BPCI primarily with a group of medical diagnosis-related groups (DRGs).OBJECTIVESIn this article, we describe our experience creating a system of response for the diverse people and diagnoses that fall into the medical DRG bundles and specifically identify organizational factors for enabling successful implementation of bundled payments.RESULTSOur experience suggests that interprofessional collaboration enabled program success.CONCLUSIONSAlthough still in its early phases, observations from our program’s strategies and tactics may provide potential insights for organizations considering engagement in the BPCI initiative.

Unplanned readmissions within 30 days after discharge: improving quality through easy prediction.

To propose an easy predictive model for the risk of rehospitalization, built from hospital administrative data, in order to prevent repeated admissions and to improve transitional care. Retrospective cohort study. Azienda Ospedaliero Universitaria Pisana (Pisa University Hospital). Patients residing in the territory of the province of Pisa (Tuscany Region) with at least one unplanned hospital admission leading to a medical Diagnosis-Related Group (DRG) in the calendar year 2012. We compared two groups of patients: patients coded as 'RA30' (readmitted within 30 days after the previous discharge) and patients coded as 'NRA30' (either admitted only once or readmitted after 30 days since the latest discharge). The effect of age, sex, length of stay, number of diagnoses, normalized number of admissions and presence of diseases on the probability of rehospitalization within 30 days after discharge was evaluated. The significant variables included in the predictive model were: age, odds ratio (OR) = 1.018, 95% confidence interval (CI) = 1.011-1.026; normalized number of admissions, OR = 1.257, CI = 1.225-1.290; number of diagnoses, OR = 1.306, CI = 1.174-1.452 and presence of cancer diagnosis, OR = 1.479, CI = 1.088-2.011. The model can be easily applied when discharging patients who have been hospitalized after an access to the Emergency Department to predict the risk of rehospitalization within 30 days. The prediction can be used to activate focused hospital-primary care transitional interventions. The model has to be validated first in order to be implemented in clinical practice.

Are Geriatricians More Efficient Than Other Physicians at Managing Inpatient Care for Elderly Patients?

To compare outcomes and measures of efficiency for hospitalized elderly adults managed by geriatricians with those managed by other physicians. Secondary data analysis using a system that integrates clinical and financial information for inpatient and outpatient services delivered throughout the University of Pittsburgh Medical Center (UPMC). Propensity scores were developed based on participant sociodemographic and clinical characteristics and used to match participants based on the attending physician's specialty (geriatrician, n = 701; nongeriatrician, n = 11,549). Multivariate analyses using generalized estimating equations methods were performed. Two UPMC hospitals in Pittsburgh, Pennsylvania. Patients aged 65 and older admitted in 2002 in a medical diagnosis-related group (DRG). Outcomes (inpatient mortality, 30-day mortality, readmission) and efficiency measures (length of stay, total costs, and surplus, which is the difference between hospital costs and payment received for an admission). Elderly adults managed by geriatricians were significantly older (P < .001) and more likely to be male (P < .001) and had more diagnoses (P < .001). Propensity scores successfully balanced characteristics managed by the two groups. Patients of geriatricians had shorter length of stay (P < .001), lower costs per admission (P < .001), and greater surplus (P < .001) with no differences in outcomes. In multivariate analyses, there were not significant differences in outcomes, but patients of geriatricians had significantly shorter length of stay and lower costs per admission and generated more surplus for the hospitals. Geriatricians were more efficient than other physicians in managing hospitalized elderly adults with medical DRGs frequently managed by geriatricians. This efficiency did not compromise patient outcomes.

Análisis de los Indicadores de Seguridad del Paciente de la Agency for Healthcare Research and Quality de los hospitales públicos de la Comunidad de Madrid

Objective To identify variability in the Patient Safety Indicators (PSI) of the Agency for Healthcare Research and Quality (AHRQ) among hospitals and by hospital level in the Madrid Health Service and to evaluate the contribution of the minimum basic data set (MBDS) records to the variability observed. Methods Data on healthcare resources were obtained from the Specialized Care Information System and from MBDS records for 2006 from 20 public and state-assisted hospitals within the Madrid Health System. AHRQ PSI were obtained from the Health Information Area of the Madrid Health Council. Data analysis included correlation coefficients and multiple linear regression. Results Surgical complications and complications due to medical or postsurgical care showed significant differences among hospitals and by hospital level. These differences were positively associated with values in some of the MBDS records. We found correlations above 0.78 between decubitus ulcer and the average weight of medical diagnosis-related groups (DRG) and between failure to rescue and average length of hospital stay and the overall average weight. At least three of the variables analyzed explained the variability in decubitus ulcer, death in low-mortality DRGs and failure to rescue. Conclusions The variability found in the Madrid Health System for the PSI is high and is associated with certain records of the MBDS. These findings should be confirmed in successive studies and in other regions, and the possible contribution of some variables of the MBDS to this variability should be assessed.

Trends in the Use of the Pulmonary Artery Catheter in the United States, 1993-2004

Although there is now substantial evidence that pulmonary artery (PA) catheterization does not reduce mortality in critically ill patients, it is unknown whether national utilization has decreased in response. To determine trends in PA catheterization use in the United States. A time trend analysis on national estimates of PA catheterization utilization from 1993-2004 using data from all US states contributing to the Nationwide Inpatient Sample. Hospital admissions for those participants aged 18 years or older were assessed, with primary analysis focused on admissions with a medical diagnosis related group and a secondary analysis focused on surgical admissions. PA catheterization was identified by 5 International Classification of Diseases, Ninth Revision procedure codes describing PA or wedge-pressure monitoring, measurement of mixed venous blood gases, or monitoring of cardiac output by oxygen consumption or other technique. Annual PA catheterization use per 1000 medical admissions. Between 1993 and 2004, PA catheterization use decreased by 65% from 5.66 to 1.99 per 1000 medical admissions (risk ratio [RR], 0.35; 95% confidence interval [CI], 0.29-0.42). Among patients who died during hospitalization, a group whose disease severity may be consistent across time, the relative decline was similar, decreasing from 54.7 to 18.1 per 1000 deaths (RR, 0.33; 95% CI, 0.28-0.38). A significant change in trend occurred following a 1996 study that suggested increased mortality with PA catheterization. The decline in utilization was similar in surgical patients (RR, 0.37; 95% CI, 0.25-0.49). Among common diagnoses associated with PA catheterization, the decline was most prominent for myocardial infarction, which decreased by 81% (RR, 0.19; 95% CI, 0.15-0.23), and least prominent for septicemia, which decreased by 54% (RR, 0.46; 95% CI, 0.38-0.54). Sensitivity analyses suggested findings were not due to artifact of changing procedure coding practice. Use of the PA catheter, previously a hallmark of critical care practice, has decreased in the United States during the last decade, possibly due to growing evidence that this invasive procedure does not reduce mortality.

A study of hospital discharges for venous thromboembolism in the south of Spain. An analysis of 19,170 cases from a regional database from 1998 to 2001

Background Our objective was to learn about the incidence of hospitalization for venous thromboembolism (VTE) in the public health care system in Andalusia and to define the profile of the patients, with special reference to the Department of Internal Medicine. Methods We analyzed the discharged data set of 32 hospitals in the Andalusian Health Care Service between 1998 and 2001, identifying the cases in whom the diagnosis was VTE. The age, sex, length of stay, outcome, number of diagnoses, diagnosis-related group (DRG), and coded procedures were studied. Results During the period studied, there were 2,228,894 discharges, 19,170 of which involved VTE. In 8494 of these, VTE was the cause of the admission. Some 3961 patients (46.6%) were admitted for pulmonary embolism (PE); 45% were discharged from internal medicine, 41% from pneumology, and 14% from other departments. The average patient age was 65, the length of stay 13.8 days, and the global in-hospital mortality rate 13%. Some 4533 cases (53.4%) were admitted due to deep vein thrombosis (DVT): 38.5% to internal medicine, 21.30% to general surgery, 12.35% to angiology, and the remainder to other departments. The length of stay was 10.6 days with an in-hospital mortality rate of 2.20%. In 7721 cases, VTE was the secondary diagnosis (after excluding 2955 cases of superficial thrombophlebitis of the upper limbs); 74% was associated with a medical DRG. Conclusions VTE is a frequent pathology in our hospitals. It shows a great variability in clinical practice although there are differences between patients treated by different specialists. VTE as secondary diagnosis was more frequent in medical inpatients.

The relationship between pediatric residency program size and inpatient illness severity and diversity.

A critical component of pediatric residency training is exposure to diverse and challenging hospitalized patients, yet little is known about the differences in pediatric inpatient educational experiences across residencies. To examine variations in inpatient illness severity and diagnostic diversity at the affiliated hospitals of small, medium, and large pediatric residencies. A retrospective analysis of hospital discharges among children aged 0 to 18 years (excluding newborns) in a sample of pediatric residency programs within the University HealthSystems Consortium. Main Outcomes of Interest The study compares the mean and median Diagnosis-Related Group (DRG) weights of hospital discharges (illness severity) as well as the percentage of discharges for the 5 most common diagnoses and the percentage of discharges for asthma (diagnostic diversity). There was no relationship between mean and median medical DRG weights and residency size (mean DRG weight: small, 0.89; medium, 0.86; and large, 0.85; small vs medium, P =.29; small vs large, P =.23). Larger programs had surgical patients with more severe illness (mean DRG weight, small, 2.11; medium, 2.08; and large, 2.47; small vs medium, P =.85; small vs large, P =.02) but less diagnostic diversity (small, 24.9%; medium, 25.9%; and large, 29.9%; small vs medium, P<.001; small vs large, P =.07). The proportion of medical discharges for asthma increased with residency size (small, 6.5%; medium, 7.4%; and large, 9.3%; small vs medium and large, P<.001). Large variations in inpatient illness severity and diagnostic diversity were seen across programs, but program size was found to be a poor indicator of inpatient learning opportunities.

Age and Laboratory Costs for Hospitalized Medical Patients

To examine the hypothesis that older hospitalized patients have higher laboratory costs than younger patients in the same severity-adjusted diagnosis-related group (DRG). We obtained hospital case mix data sets (1995-1997) from the Massachusetts Division of Health Care Finance and Policy. We selected discharge abstracts from 4 medical DRGs, at 5 large academic hospitals (n = 15,265) and 5 midsized community hospitals (n = 10,540), for analysis. We converted laboratory and blood product charges to direct costs using the department-specific ratio of cost to charges. We adjusted diagnostic groups for severity of comorbid conditions and complications using the refined DRG method. Hospital length of stay (LOS), laboratory direct cost (LDC) per hospitalization, LDC per hospital day, and ratio of LDC to total direct cost. Hospital LOS was longer for older patients in all comparisons. Laboratory direct cost per hospitalization was higher for older patients in some DRGs, but lower in other DRGs. Laboratory direct cost per hospital day was almost always less for older patients than for younger patients, both at academic and community hospitals. Data stratification by gender, admission status, and principal diagnosis yielded substantially the same pattern of cost differences observed within the larger data set. Older medical patients have longer hospital stays and generally higher costs. These patients also have a significantly decreased rate of laboratory resource consumption over the course of hospitalization (LDC per hospital day), as well as lower laboratory costs as a proportion of total costs. Age-specific differences in LOS and cost parameters were essentially unchanged after controlling for several potential sources of bias.

Geographic variation in the use of post-acute care.

To assess the extent and consistency of geographic differences in the use of post-acute care (PAC), and the stability of this pattern of variation. The 5 percent Medicare data sample for 1996, 1997, and the first eight months of 1998 were used. Patterns of PAC use for various Diagnosis-related Groups (DRGs) cross states (33 with enough cases per year) and census divisions were examined. The consistency of relative rankings for overall PAC use and use within defined DRGs was compared. PAC use varied substantially across regions. For example, the extent of any PAC use for stroke patients varied by 12 percentage points among census regions in 1998. The pattern of PAC use was quite consistent across years; 30 of the 36 possible Spearman rank order correlations were statistically significant with coefficients ranging from 0.35 to 0.95 among the DRGs studied. The correlations among DRGs were generally high. For skilled nursing facility use, all the correlations were above 0.5 and were statistically significant; in general the patterns were highest within medical DRGs (0.65-0.93). The variation in PAC use is not a statistical artifact. It is likely the result of several forces: practice styles, supply of services, and local regulatory practices.

Les groupes homogènes de malades proposés pour la réanimation. Étude de la pertinence de cette classification basée sur un calcul des coûts

Aim: The aim of this study was to determine the real cost of a patient's stay in an intensive care unit (ICU), to analyze the variance within 16 new specific diagnosis related groups (DRG) proposed by the Société de Réanimation de Langue Française, and thereby to assess the relevance of this classification. Method: Real costs were determined daily and on a duration of stay basis for each DRG using a computerized clinical information system (Carevue TM , Hewlett Packard) and customized software. The prospect cost assessment was made after two data processing steps: conversion of flowsheet data into 15 cost-category units, and multiplication of the number of cost-category units by set unit price.The validity of the calculated costs was assessed by comparing them with actual ICU costs obtained from the administration. Study setting: This study was carried out in a general ICU with 22 beds in a teaching hospital with a total of 1,398 short-stay beds. Patients: This prospective study was based on hospital admissions over a 1-year period, i.e., consecutive hospitalizations between January 1st and December 31st. Results: An analysis was.made of 859 ICU stays, corresponding to 7,142 days of hospitalization. The results were as follows: mean SAPS II: 40.9 ± 18, overall ICU mortality: 25%. During the study period, the total calculated costs per patient accounted for 93.4% of the ICU expenditure allocated by the administration. For each DRG in the ICU, within-group homogeneity was observed, and the coefficient of variationfor costs (SD/mean value) was consistently less than 1. The 16 DRG explained 60.4% of the cost variance. SAPS II had ho influence on total cost prediction. Conclusions: This analysis confirms the reliability of the results that have already been obtained in another: study, The cost analysis homogeneity obtained bythis classification was superior to that of the majority of medical DRGs. In this study, the SAPS II did not play a significant role in predicting the cost of the hospital stay. Buts de l’étude: Mesurer le coût réel des séjours en réanimation et analyser leur variance à l’intérieur des 16 nouveaux groupes homogènes de malade (GHM) proposés par la Société de réanimation de langue française, afin d’évaluer la pertinence de cette classification. Protocole: L’évaluation prospective des coûts a été réalisée à partir des données d’un dossier de soins informatisé (Carevue TM , Hewlett Packard), à l’aide d’un logiciel spécifique, après deux étapes de traitement de l’information: conversion des données en 15 unités d’œuvre, multiplication du nombre d’unités d’œuvre par leur prix unitaire au tarif de l’établissement. La validité des coûts a été évaluée par rapport à la comptabilité analytique de l’hôpital. Les coûts ont été calculés par jour et par séjour pour chaque GHM. Lieu de l’étude: Un service de réanimation polyvalente de 22 lits faisant partie d’un hôpital universitaire de 1 398 lits de court séjour. Sélection des patients: Ce travail prospectif a porté sur les admissions d’une année. Les séjours des patients admis à partir du 1 er janvier 1998 et sortis avant te 1 er janvier 1999 ont été sélectionnés. Résultats: Huit cent cinquante-neuf séjours correspondant à 7142 journées, IGS II; 40,9 ± 18, mortalité: 25%. Pour la période de l’étude, le cumul des coûts calculés par patient avec cette méthode représente 93,4% des coûts établis par la comptabilité analytique de l’hôpital. Pourchaque GHM de réanimation, le coefficient de variation (écart-type [SD]/moyenne) est toujours inférieur à 1. Les 16 GHM expliquent 60,4 % de la variance des coûts. Conclusions: Cette étude renforce la crédibilité des résultats déjà obtenus dans une autre étude. L’homogénéité des coûts obtenue par cette classification est supérieure à celle de la plupart des GHM médicaux. Dans cette étude, l’IGS 2 apporte une part négligeable dans la prédiction du coût d’un séjour.

Identifying poor-quality hospitals. Can hospital mortality rates detect quality problems for medical diagnoses?

Many groups involved in health care are very interested in using external quality indices, such as risk-adjusted mortality rates, to examine hospital quality. The authors evaluated the feasibility of using mortality rates for medical diagnoses to identify poor-quality hospitals. The Monte Carlo simulation model was used to examine whether mortality rates could distinguish 172 average-quality hospitals from 19 poor-quality hospitals (5% versus 25% of deaths being preventable, respectively), using the largest diagnosis-related groups (DRGs) for cardiac, gastrointestinal, cerebrovascular, and pulmonary diseases as well as an aggregate of all medical DRGs. Discharge counts and observed death rates for all 191 Michigan hospitals were obtained from the Michigan Inpatient Database. Positive predictive value (PPV), sensitivity, and area under the receiver operating characteristic curve were calculated for mortality outlier status as an indicator of poor-quality hospitals. Sensitivity analysis was performed under varying assumptions about the time period of evaluation, quality differences between hospitals, and unmeasured variability in hospital casemix. For individual DRG groups, mortality rates were a poor measure of quality, even using the optimistic assumption of perfect casemix adjustment. For acute myocardial infarction, high mortality rate outlier status (using 2 years of data and a 0.05 probability cutoff) had a PPV of only 24%, thus, more than three fourths of those labeled poor-quality hospitals (high mortality rate outliers) actually would have average quality. If we aggregate all medical DRGs and continue to assume very large quality differences and perfect casemix adjustment, the sensitivity for detecting poor-quality hospitals is 35% and PPV is 52%. Even for this extreme case, the PPV is very sensitive to introduction of small amounts of unmeasured casemix differences between hospitals. Although they may be useful for some surgical diagnoses, DRG-specific hospital mortality rates probably cannot accurately detect poor-quality outliers for medical diagnoses. Even collapsing to all medical DRGs, hospital mortality rates seem unlikely to be accurate predictors of poor quality, and punitive measures based on high mortality rates frequently would penalize good or average hospitals.

Can Early Re-Admission Rates Accurately Detect Poor-Quality Hospitals?

There is widespread interest in using external quality measures, such as early re-admission rates (ERRs), to evaluate hospital quality. To evaluate the feasibility of using ERRs to identify poor-quality hospitals, a Monte Carlo simulation model was developed that describes the predictive power of ERRs for the 190 hospitals in Michigan using different assumptions concerning the distribution and variability of quality problems, the number of years of data aggregated, and unmeasured case-mix differences. The ability of ERRs to distinguish 171 average-quality hospitals from 19 poor-quality hospitals (assigned to have 5% vs. 15% premature discharges) was evaluated. First, the largest diagnosis-related groups (DRGs) were studied to determine if they included cardiac, gastrointestinal, pulmonary, and neurologic diseases. Despite using the highly optimistic assumptions that premature discharges are readmitted 50% more frequently than appropriately timed discharges and that no ERR variation was caused by unmeasured case-mix differences between hospitals, the results were poor. For example, for DRG 127 (heart failure), high ERR outlier status (using a .05 probability cutoff) had a positive predictive value of only 36%, meaning that approximately two thirds of hospitals labeled "poor-quality" (high ERR outliers) were false-positive results. Next, we repeated the simulation with sample sizes aggregated for all medical DRGs. The positive predictive value was 72%, but was very sensitive to ERR variability due to non-quality-related factors (e.g., unmeasured case mix). The positive predictive value decreases to 45% if unmeasured case mix accounts for even 10% of observed hospital ERR variation. The circumstances under which DRG-specific ERRs would be useful to detect poor-quality hospitals are unlikely to occur. Even collapsing to all medical DRGs, ERRs are likely to be accurate predictors only if quality differences are quite large and if unmeasured case-mix differences account for a small amount of interhospital variation in ERRs.

Increased health care costs associated with ED overcrowding

The overcrowding of emergency departments (EDs) with inpatients results in an increased average inpatient length of stay; therefore, overcrowded hospitals have increased costs per patient. All admissions through the ED to our institution for 1988, 1989, and 1990 were reviewed. These admissions were analyzed based on whether they had spent less than 1 day or more than 1 day in the ED, after they had been admitted to the hospital and were waiting for a bed assignment. Analyses were performed for the five medical diagnosis-related groups, with the highest volumes of admissions via the ED. All categories were reviewed on the basis of whether or not the payor was Medicare. This was a retrospective data analysis of 3 years worth of hospital and ED length of stay. There was no intervention. The total number of patients admitted via the ED for 1988, 1989, and 1990 was 26,020. In 1988, 19% of admissions via the ED spent more than 1 day in the ED. The total hospital length of stay for this 19% was 11% longer than for the group who reached an inpatient bed on the first hospital day. In 1989, 32% of admissions via the ED remained in the ED for more than 1 day and had a 13% increase in total hospital length of stay. In 1990, 25% of admissions via the ED spent more than 1 day in the ED and had a 10% increase in total hospital length of stay. Inpatients who remained in the ED after admission had a greater average length of stay than those who were promptly transferred to inpatient units. This increased length of stay means an increased cost per patient; the overall impact for this 490-bed hospital was $6.8 million during the 3-year study period.

DRGs and Australian Psychiatry

The introduction of diagnosis related groups (DRGs) as a basis for funding in the United States has revealed several shortcomings in current DRG systems. Overall, DRGs have proven to be poor predictors of cost, accounting for approximately one third of the variation in cost and length of stay for surgical DRGs and falling to less than 10% for medical DRGs. Their ability to contain costs also remains uncertain, with savings associated with reductions in length of stay being offset by increased readmission rates. Given the increasing commitment of government to casemix approaches to funding it is suggested that psychiatry should participate in the process of solving the problems so far identified with DRG systems. Participation would, it is hoped, create a focussed debate about the provision of a "gold standard" of care for all patients. The evaluation and refinement of existing DRGs is urgently needed and could utilise a number of comprehensive data bases which already exist across the country. Alternatives to diagnosis such as functional status and treatment needs should also be explored.

Watching the doctor-watchers. How well do peer review organization methods detect hospital care quality problems?

To determine how well one state's peer review organization (PRO) judged the quality of hospital care compared with an independent, credible judgment of quality of care. Retrospective study comparing a PRO's review, including initial screening, physician review, and final judgments, with an independent "study judgment" based on blinded, structured, implicit reviews of hospital records. One state's medical and surgical Medicare hospitalizations during 1985 through 1987 audited randomly by the state's PRO. Stratified random sampling of records: 62 records that passed the PRO initial screening process and were not referred for PRO physician review; 50 records that failed the PRO screen and then were confirmed by PRO physicians to be "quality problems." None. A study judgment of below standard or standard or above based on the mean of overall ratings by five internists for records in medical diagnosis related groups (DRGs) and by five internists and five surgeons for surgical DRGs. Each step in the PRO review was evaluated for how many records passing or failing that step were judged standard or above or below standard in the study (positive and negative predictive value) and how well that step classified records that the study judged below standard or standard or above (sensitivity and specificity). An estimated 18% of records reviewed by the PRO were below standard according to the study judgment, compared with 6.3% quality problems according to the PRO's final judgment (difference, 12%; 95% confidence interval, 1 to 23). The PRO's initial screening process failed to detect and refer for PRO physician review two of three records that the study judged below standard. In addition, only one of three of the records that PRO physicians judged to be quality problems were judged below standard by the study judgment. Therefore, the PRO's final quality of care judgment and the study judgment agreed little more than expected by chance, especially about poor quality of care. Although the PRO correctly classified 95% of the records that the study judged standard or above, it detected only 11% of records judged below standard by the study. Most of all, this PRO review process would be improved by additional preliminary screens to identify the 67% of records that the study judged below standard but that passed its initial screening. The screening process also must be more accurate in order to be cost-effective, as it was only slightly better than random sampling at correctly identifying below standard care. More reproducible physician review is also needed and might be accomplished through improved reviewer selection and training, a structured review method, and more physician reviewers per record.

Watching the Doctor-Watchers

<h3>Objective.</h3> —To determine how well one state's peer review organization (PRO) judged the quality of hospital care compared with an independent, credible judgment of quality of care. <h3>Design.</h3> —Retrospective study comparing a PRO's review, including initial screening, physician review, and final judgments, with an independent "study judgment" based on blinded, structured, implicit reviews of hospital records. <h3>Setting.</h3> —One state's medical and surgical Medicare hospitalizations during 1985 through 1987 audited randomly by the state's PRO. <h3>Sample.</h3> —Stratified random sampling of records: 62 records that passed the PRO initial screening process and were not referred for PRO physician review; 50 records that failed PRO screen and then were confirmed by PRO physicians to be "quality problems." <h3>Intervention.</h3> —None. <h3>Main Outcome Measure.</h3> —A study judgment of below standard or standard or above based on the mean of overall ratings by five internists for records in medical diagnosis related groups (DRGs) and by five internists and five surgeons for surgical DRGs. Each step in the PRO review was evaluated for how many records passing or failing that step were judged standard or above or below standard in the study (positive and negative predictive value) and how well that step classified records that the study judged below standard or standard or above (sensitivity and specificity). <h3>Results.</h3> —An estimated 18% of records reviewed by the PRO were below standard according to the study judgment, compared with 6.3% quality problems according to the PRO's final judgment (difference, 12%; 95% confidence interval, 1 to 23). The PRO's initial screening process failed to detect and refer for PRO physician review two of three records that the study judged below standard. In addition, only one of three of the records that PRO physicians judged to be quality problems were judged below standard by the study judgment. Therefore, the PRO's final quality of care judgment and the study judgment agreed little more than expected by chance, especially about poor quality of care. Although the PRO correctly classified 95% of the records that the study judged standard or above, it detected only 11% of records judged below standard by the study. <h3>Conclusions.</h3> —Most of all, this PRO review process would be improved by additional preliminary screens to identify the 67% of records that the study judged below standard but that passed its initial screening. The screening process also must be more accurate in order to be cost-effective, as it was only slightly better than random sampling at correctly identifying below standard care. More reproducible physician review is also needed and might be accomplished through improved reviewer selection and training, a structured review method, and more physician reviewers per record. (<i>JAMA</i>. 1992;267:2349-2354)