Preliminary validation of the Review of Musculoskeletal System (ROMS) questionnaire.

BACKGROUND: Measurement of clinical outcomes is necessary to define best practice. It requires a validated tool that can be easily applied as part of clinical practice. We present the preliminary validation of a brief self-reported Review of Musculoskeletal System (ROMS) questionnaire that captures functional limitations due to musculoskeletal problems and other medical and emotional conditions. METHODS: Data were derived from a clinical outcomes database (Orthopaedic Minimal Data Set [OrthoMiDaS]) that combines patient-reported data collected as part of routine care and secondary data extracted from electronic medical records. The study utilized 82,873 encounters collected from 24,116 consecutive patients with problems in the upper and lower extremities. In addition to the ROMS, the study used version 2 of the Short Form-12 (SF-12v2), the Penn Shoulder Score (PSS), the Hip disability and Osteoarthritis Outcome Score (HOOS), and the Knee injury and Osteoarthritis Outcome Score (KOOS) questionnaires. Fifteen cross-sectional samples were used to evaluate the floor and ceiling effects as well as the construct and content validity. Five longitudinal cohorts were used to measure test-retest reliability and responsiveness. Standard statistical tests were applied. RESULTS: The floor and ceiling effects of the ROMS questionnaire in patients with shoulder, hip, and knee problems ranged from 1.3% to 8.5%. Construct-validity tests confirmed convergent and divergent validity of the ROMS. The tests also justified its additional value when the ROMS was used with joint-specific tools. When measuring test-retest reliability of the ROMS scales, intraclass correlation ranged from 0.80 to 0.90 at approximately one week and from 0.71 to 0.87 at approximately four weeks. Responsiveness of the ROMS was greater than that of the SF-12 and less than that of the joint-specific questionnaires. CONCLUSIONS: The ROMS is compatible with routine clinical process and has good psychometric properties in patients with shoulder, hip, and knee disorders. It can be used as a primary outcome tool for large observational studies and can supplement more specific tools in controlled studies. CLINICAL RELEVANCE: The ROMS was developed as a tool to measure and monitor the clinical status of the musculoskeletal system in a population of patients during and after treatment as well as over time.

Effect of surgeon experience on femoral component size selection during total knee arthroplasty.

Femoral component size selection during total knee arthroplasty should not vary from surgeon to surgeon for patients with the same bone size. This study explored if systematic variations in femoral component size selection exist. Thirteen surgeons' choices of femoral component size (Duracon, n = 1388; Triathlon, n = 740) were analyzed using a generalized linear model with femoral component size as the dependent variable and surgeon identification, years in practice, and adult reconstruction fellowship training as the independent variables. The model adjusted for differences in bone size. It was found that more experienced surgeons implant larger femoral components. New instruments and training protocols may be necessary to adjust for surgeon experience.

Prognosticators in thigh soft tissue sarcomas.

BACKGROUND: In sarcoma patients the roles of smoking history, family cancer history, and leukoreduced blood transfusions have not been studied and the effect of preoperative radiation on blood loss has not been examined. METHODS: Seventy-seven patients with non-metastatic and non-recurrent thigh sarcomas surgically treated at the Cleveland Clinic were identified. Among patient variables studied were: close family history of cancer, perioperative transfusion history, smoking history, and radiation history. Median follow-up for the survivors was 3.2 years. RESULTS: We found that tumor grade, transfusion >3 U (P = 0.022), and pre- or post-operative radiation therapy (P = 0.041) were risk factors for distant metastasis. Tumor grade (P = 0.008), positive smoking history (P = 0.039), and >3 U of non-leukoreduced blood transfused (P = 0.037) were risk factors for death of any-cause. Close family history of cancer correlated with having a grade 3 sarcoma (P = 0.044). Neoadjuvant radiotherapy correlated with >3 U of blood transfused (P = 0.001) and biopsy performed at the treating institution led to a significant decrease in rate of recurrence (P = 0.016). CONCLUSIONS: We present novel findings in terms of transfusions, family history of cancer and site of initial biopsy in sarcoma patients.

The effect of medical comorbidity on self-reported shoulder-specific health related quality of life in patients with shoulder disease.

BACKGROUND: The purpose of this study was to investigate further the effect of medical comorbidity on a patient reported shoulder specific health related quality of life (HRQoL) measure. We investigated which types of comorbidities have a detrimental effect upon shoulder specific HRQoL. We hypothesized that general medical comorbidity would not negatively affect shoulder specific HRQoL questionnaires, but that comorbidities specific to the chest region would, when properly controlling for other patient factors. METHODS: A cohort of 173 consecutive patients who underwent shoulder surgery for osteoarthritis and/or rotator cuff repair was extracted from a clinical outcomes database. Their health related quality of life (HRQoL) was evaluated with the University of Pennsylvania (PENN) shoulder score and the Short Form-36 (SF-36). Nonadjusted and multivariate risk-adjusted models were built to investigate the effect of medial comorbidity on shoulder specific HRQoL and were tested using linear modeling. RESULTS: Nonadjusted models showed patients with more total comorbidities (P=.01) and more chest-related comorbidities (P=.006) had lower PENN scores. But, when risk adjusting for other patient factors, the PENN scores decreased with an increase in the number of chest comorbidities (P=.008), but not the number of total comorbidites (P=.391) or other (nonchest) comorbidities (P=.163). CONCLUSION: Shoulder specific HRQoL measures are joint specific, but they are influenced by disease or conditions that affect the chest region. This may be important in understanding why patients with certain comorbid diseases report worse shoulder pain and function and may respond differently to treatment over time.

Factors affecting recovery after arthroscopic labral debridement of the hip.

PURPOSE: The purpose of this study was to develop and validate a model predicting whether patients would have shorter-than-typical or longer-than-typical recoveries after hip arthroscopy for labral tears. METHODS: We retrospectively reviewed 268 cases of hip arthroscopy implemented between 2000 and 2007 by 2 orthopaedic surgeons at our institution. The development cohort consisted of patients with magnetic resonance angiography-identified labral tears and a history and physical examination consistent with either labral pathology or loose bodies. Univariate analysis targeted preoperative patient characteristics correlated with the risk of longer-than-typical recoveries. Multivariate logistic regression was applied to generate an algorithm predicting risk of longer-than-typical recovery based on baseline characteristics. The algorithm was tested in the validation sample of 52 patients who were treated in 2007 and was found to be valid. RESULTS: Five predictors for longer-than-typical recovery were identified: Workers' Compensation status, female gender, use of pain medications, presence of a limp, and presence of a lateral labral tear. The multivariate algorithm was developed and successfully validated. CONCLUSIONS: This study identifies many new predictors of recovery, and it also corroborates those that have already been identified. The 5 predictors for longer-than-typical recovery identified by our validated multivariate algorithm were Workers' Compensation status, female gender, use of pain medications, presence of a limp, and presence of a lateral labral tear. LEVEL OF EVIDENCE: Level IV, therapeutic case series.

Patterns of functional improvement after revision knee arthroplasty.

BACKGROUND: Despite the increase in the number of total knee arthroplasty revisions, outcomes of such surgery and their correlates are poorly understood. The aim of this study was to characterize patterns of functional improvement after revision total knee arthroplasty over a two-year period and to investigate factors that affect such improvement patterns. METHODS: Three hundred and eight patients in need of revision surgery were enrolled into the study, conducted at seventeen centers, and 221 (71.8%) were followed for two years. Short Form-36 (SF-36), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and Lower-Extremity Activity Scale (LEAS) scores were collected at baseline and every six months for two years postoperatively. A piecewise general linear mixed model, which models correlation between repeated measures and estimates separate slopes for different follow-up time periods, was employed to examine functional improvement patterns. RESULTS: Separate regression slopes were estimated for the zero to twelve-month and the twelve to twenty-four-month periods. The slopes for zero to twelve months showed significant improvement in all measures in the first year. The slopes for twelve to twenty-four months showed deterioration in the scores of the WOMAC pain subscale (slope = 0.67 +/- 0.21, p < 0.01) and function subscale (slope = 1.66 +/- 0.63, p < 0.05), whereas the slopes of the other measures had plateaued. A higher number of comorbidities was consistently the strongest deterrent of functional improvement across measures. The modes of failure of the primary total knee arthroplasty were instrument-specific predictors of outcome (for example, tibial bone lysis affected only the SF-36 physical component score [coefficient = -5.46 +/- 1.91, p < 0.01], while malalignment affected both the SF-36 physical component score [coefficient = 5.41 +/- 2.35, p < 0.05] and the LEAS score [coefficient = 1.42 +/- 0.69, p < 0.05]). Factors related to the surgical technique did not predict outcomes. CONCLUSIONS: The onset of worsening pain and knee-specific function in the second year following revision total knee arthroplasty indicates the need to closely monitor patients, irrespective of the mode of failure of the primary procedure or the surgical technique for the revision. This information may be especially important for patients with multiple comorbidities.

Functional outcome of arthroscopic rotator cuff repairs: a correlation of anatomic and clinical results.

Even though several studies have revealed excellent clinical results with arthroscopic repair of rotator cuff tendons, poor healing of the repair and retearing of the tendon occur in many cases. Patterns of outcome correlating functional capacity and anatomic integrity of the repaired rotator cuff are not well defined. The goal of this prospective study was to determine the pattern of anatomic and functional outcomes among patients undergoing single-row arthroscopic rotator cuff repair. This study confirmed that single-row arthroscopic repair of small- and medium-sized supraspinatus tendon tears significantly improves rotator cuff integrity and functional outcomes. A completely healed tendon was observed in 60% of the cases. Age is a predictor of cuff integrity after the operation. Functional improvement was greater and significant in patients with complete healing at follow-up; however, a recurrent tear did not preclude positive functional results.

Functional improvement after total knee arthroplasty revision: new observations on the dimensional nature of outcome.

BACKGROUND: Despite the numerous outcomes measures described it remains unclear what aspects of patient outcome are important in determining actual improvement following total knee arthroplasty revisions (TKAR). We performed a prospective cohort study of TKAR to determine the components of clinical improvement and how they are related and best measured. METHODS: An improvement scale was devised utilizing data from 186 consecutive TKAR patients on SF-36 physical (PCS) and mental (MCS) components, Western Ontario and McMaster Universities Osteoarthritis (WOMAC) Index, Knee Society Score (KSS), a novel Activity Scale (AS) and a physician derived severity assessment scale performed both preoperatively and at 6 month post-operative follow-up. The change in each of these scores was analyzed using factor analysis, deriving a composite improvement scale. RESULTS: All the instruments demonstrated statistically significantly better scores following TKAR (except the SF-36 MCS). Furthermore, all significant correlations between the scores were positive. Statistical factor analysis demonstrated that scores could be arranged into 4 related factor groupings with high internal consistency (Cronbach Alpha = 0.7). Factor 1 reflected patient perceived functional outcomes, Factor 2 activity levels, Factor 3 the MCS and Factor 4 the KSS. CONCLUSION: This study demonstrates that improvement following TKAR has a multidimensional structure. The improvement scales represent a more coordinated method of the previously fragmented analysis of TKAR outcomes. This will improve assessment of the actual effectiveness of TKAR for patients and what aspects of improvement are most critical.

Efficacy of growth factors compared to other therapies for low-risk myelodysplastic syndromes.

The myelodysplastic syndromes (MDS) represent a heterogeneous group of disorders. Low-risk MDS represent a subgroup with a relatively good prognosis, but with few trials evaluating outcomes. A pooled analysis based upon a MEDLINE search identified 162 original articles describing patient characteristics and effect of therapy on 2592 individuals with pathologically confirmed refractory anaemia or refractory anaemia with ringed sideroblasts with <5% bone marrow blasts. Treatments were categorised as growth factors (GF) or non-growth factors (NGF). International Prognostic Scoring System (IPSS) score was documented or calculated when possible. Responses and outcomes were standardised according to the International Working Group MDS criteria. Growth factors produced higher overall response rates (39.5% vs. 31.4% for NGF, P = 0.019), while NGF yielded better CR/PR rates (25.6% vs. 9.1% for GF, P = 0.03). Over 2 years of follow-up, those receiving GF demonstrated greater overall and progression-free survival than NGF, after controlling for baseline patient characteristics. Decision tools need to be developed to determine which therapy to choose for patients with low-risk MDS.

Who recommends long-term care matters.

PURPOSE: Making good consumer decisions requires having good information. This study compared long-term-care recommendations among various types of health professionals. DESIGN AND METHODS: We gave randomly varied scenarios to a convenience national sample of 211 professionals from varying disciplines and work locations. For each scenario, we asked the professional to recommend the appropriate forms of long-term care. RESULTS: Although the professional respondents used the full spectrum of options offered to them, some professionals tended to favor the sector they worked in. Advanced practice nurses recommended day care and homemaking more and adult foster care less. Gerontologists used skilled nursing-facility placement more actively and rehabilitation, homemaking, and home health care less actively. Geriatricians and primary care physicians both favored rehabilitation and skilled nursing-facility care and were both less enthusiastic about assisted living, homemaking, and informal care, but the geriatricians favored day care more than did the primary care physicians. Registered nurses were highly supportive of assisted living, adult foster care, homemaking, and home health care, and they opposed skilled nursing-facility care. Social workers were less likely than other participants to endorse rehabilitation and adult foster care. IMPLICATIONS: Because consumer preference should be a major factor in making long-term-care decisions, many consumers need information about what options may best fit their situation. In the absence of empirical data on which types of long-term care work best for whom, consumers have to rely on expert judgment-but that judgment varies. Clients should be aware that an expert's background (as defined by discipline and work situation) may affect his or her recommendations. Each discipline appears to have its own set of experiences and beliefs that may influence recommendations.

Variations on a theme called PACE.

BACKGROUND: The Program for All-inclusive Care of the Elderly (PACE) has been hailed as successful but of limited appeal. This study contrasts the effects on hospital utilization of PACE and a more liberal variant, the Wisconsin Partnership Program (WPP). METHODS: Hospital and emergency room (ER) utilization data from two sites that used both PACE and WPP to serve elderly clients were compared. The analysis of utilization was conducted using a cross-sectional longitudinal approach. The statistical significance of the difference between WPP and PACE groups was calculated by using regressions that adjusted for gender, race (white/nonwhite), age, original reason for entitlement in Medicare (elderly/disabled), dual eligibility, diagnoses during the previous 6 months, and county of residence. RESULTS: The PACE enrollees had fewer hospital admissions, preventable hospital admissions, hospital days, ER visits, and preventable ER visits than the WPP enrollees had. There was no difference in the length of hospital stays. CONCLUSIONS: PACE is more effective in controlling hospital and ER utilization than is the more flexible variant (WPP).

Predicting nursing facility residents' quality of life using external indicators.

PURPOSE: A newly developed brief measure of nursing facility (NF) resident self-reported quality of life (QOL) has been proposed for inclusion in a modified version of the minimum data set (MDS). There is considerable interest in determining whether it is possible to develop indicators of QOL that are more convenient and less expensive than direct, in-person interviews with residents. DESIGN AND METHODS: QOL interview data from 2,829 residents living in 101 NFs using a 14-item version of a longer instrument were merged with data from the MDS and the Online Survey and Certification Automated Record (OSCAR). Bivariate and multivariate hierarchical linear modeling were used to assess the association of QOL with potential resident and facility level indicators. RESULTS: Resident and facility level indicators were associated with self-reported QOL in the expected direction. At the individual resident level, QOL is negatively associated with physical function, visual acuity, continence, being bedfast, depression, conflict in relationships, and positively associated with social engagement. At the facility level, QOL is negatively associated with citations for failing to accommodate resident needs or providing a clean, safe environment. The ratio of activities staff to residents is positively associated with QOL. This study did not find an association between QOL and either use of restraints or nurse staff levels. Approximately 9 percent of the total variance in self-reported QOL can be attributed to differences among facilities; 91 percent can be attributed to differences among residents. Resident level indicators explained about 4 percent of the variance attributable to differences among residents, and facility factors explained 49 percent of the variance attributable to differences among NFs. However, the different variables explained only 10 percent of the variance in self-reported QOL. IMPLICATIONS: A brief self-report measure of NF resident QOL is consistently associated with measures that can be constructed from extant data sources. However, the level of prediction possible from these data sources does not justify reliance on external indicators of resident QOL for policy purposes.

The effects of a variant of the program for all-inclusive care of the elderly on hospital utilization and outcomes.

OBJECTIVES: To compare the effects of the Wisconsin Partnership Program (WPP) on hospital, emergency department (ED), and nursing home utilization with those of traditional care. DESIGN: Quasi-experimental longitudinal cohort design. SETTING: Selected counties in Wisconsin. PARTICIPANTS: WPP elderly enrollees and two matched control groups consisting of frail older people enrolled in fee-for-service insurance plans, Medicare, and Medicaid and receiving home- and community-based waiver services, one from the same geographic area as the WPP and another from a location in the state where the WPP was not offered. MEASUREMENTS: Data came from administrative records. Regression and survival analyses were adjusted for case-mix variables. RESULTS: No significant differences in hospital utilization, ED visits, preventable hospitalizations, risk of entry into nursing homes, or mortality were found. WPP enrollees had more contact with care providers than did controls. CONCLUSION: WPP did not dramatically alter the pattern of care. Part of the weak effect may be attributable to the small numbers of WPP cases per participating physician.

Hormonal therapy: ACL and ankle injury.

BACKGROUND: No definitive explanation for the difference in rate of male and female noncontact ACL injury has been found. The hormonal environment, known to be different in men and women has been hypothesized as a possible source for this difference in injury rate. PURPOSE: To confirm earlier work looking at periodicity of noncontact ACL injury. To increase sample size by adding ankle sprains. To determine the rate of noncontact ACL injury and ankle sprains in collegiate basketball and soccer. To determine if the use of oral contraceptives affects the rate of noncontact ACL injury and ankle sprains. METHODS: Data was collected from a sample of NCAA schools over the 2000-2001 basketball and the 2001-2002 basketball and soccer seasons. RESULTS: Recall and prospective data collection of length of menstrual cycle did not produce equivalent results. Periodicity was present only in the recall group of "off pill" users. The rate of noncontact ACL injury and noncontact ankle sprains was twice as high in basketball as in soccer. There was no difference in rate of injuries between those athletes using hormonal therapy and those athletes not using hormonal therapy. CONCLUSIONS: Noncontact ACL injuries and ankle sprains occurred at significantly higher rates in basketball than in soccer but this rate difference was not linked to hormonal therapy usage. The overall rate of noncontact ACL injury and ankle sprain to women's collegiate basketball and soccer players is very low.

Proxy sources for information on nursing home residents' quality of life.

OBJECTIVES: This study explores how well staff and family proxies' reports on selected quality-of-life (QOL) domains (comfort, dignity, functional competence, privacy, meaningful activity, food enjoyment, relationships, security, and autonomy) correspond to residents' own reports. METHODS: We compared QOL domain scores for nursing home residents and 1,326 staff proxies and 989 family proxies at the individual and facility level using means, Pearson correlation statistics, and intraclass correlations. Regression models adjusted for residents' age, gender, length of stay, ability to perform activities of daily living, and cognition. RESULTS: For each domain in more than half the cases, proxy means were within 1 SD of the resident means. Resident and family proxy individual reports for selected domains were correlated at 0.14 to 0.46 (all p <.000). Resident and staff proxy individual reports were correlated at 0.13 to 0.37 (all p <.000). Correlation of mean levels by facility for staff proxies was 0.26 to 0.64 (generally p <.05) and for family proxies 0.13 to 0.61 (p <.01 except for one domain). DISCUSSION: Although staff and family proxy domain scores are significantly correlated with resident scores, the level of correlation suggests they cannot simply be substituted for resident reports of QOL. Determining how proxy reports can be used for residents who cannot be interviewed at all remains an unresolved challenge.

Development and validation of a lower-extremity activity scale. Use for patients treated with revision total knee arthroplasty.

BACKGROUND: Valid outcome measurement tools are required to reliably demonstrate the effectiveness and clinical outcomes of lower-extremity arthroplasty. Having ascertained a lack of a practical and valid measure of the change in actual daily physical activity that occurs prior to and following lower-limb arthroplasty, we developed and validated a lower-extremity activity scale. METHODS: The eighteen-level self-administered scale was developed with the aid of content experts to ensure face validity. Validity and reliability were assessed with the use of (1) pedometer measurements of seventy subjects over seven days; (2) next-of-kin proxy measurements of the activity levels of ninety patients before they underwent lower-limb arthroplasty; and (3) application, and correlation with the Western Ontario and McMaster Universities Osteoarthritis Index scores, in a prospective seventeen-center clinical study of 297 consecutive patients undergoing revision total knee arthroplasty. In this latter study, demographic and comorbidity data were also collected. Univariate and bivariate correlations were performed, and a multivariate structured equation modeling approach was used to further test responsiveness, reliability, and validity of the lower-extremity activity scale. RESULTS: Pedometer readings correlated with the activity levels derived with the lower-extremity activity scale (r = 0.79). Of note was the finding that age, weight, and body mass index did not correlate well with the average number of steps per day (r = -0.32, -0.32, and -0.25, respectively). A significant correlation was found between the lower-extremity activity scores recorded by the patients and those reported by their next of kin (Pearson correlation, r = 0.715; p = 0.0001) and between the initial lower-extremity activity scores and two-week-retest scores (intraclass correlation = 0.9147; p < 0.0001), demonstrating the validity and reliability of the scale. The lower-extremity activity scale was responsive, accurately reflecting changes in the patient's condition between baseline and the time of follow-up (p < 0.001), and it was reliable, with baseline values correlating with follow-up scores (p < 0.001). The convergent validity of the lower-extremity activity scale was established by correlations with the function scores (r = -0.301, p < 0.001) and pain scores (r = -0.241, p < 0.001) derived with the Western Ontario and McMaster Universities Osteoarthritis Index and with a higher number of comorbidities (r = -0.244, p < 0.001). Multivariate path modeling further demonstrated diminished activity in patients who had more difficulty in functioning and a greater number of comorbidities. CONCLUSIONS: We developed a lower-extremity activity scale and validated that it was an effective instrument for the assessment of patients' actual activity levels. It is easy to apply and interpret, and it is valid and ready for use in the clinical setting. This scale will allow more accurate analysis and prediction of outcomes. Consequently, it will become a useful, practical adjunct to objective clinical decision-making and intervention for patients undergoing arthroplasty.

The functional outcomes of total knee arthroplasty.

BACKGROUND: As an elective procedure, total knee arthroplasty is under scrutiny to evaluate its cost-effectiveness. In this review, we examined the available literature on total knee arthroplasty to assess the evidence regarding factors associated with better functional outcomes. METHODS: A structured literature search of English-language databases was performed to identify studies of the functional outcomes of total knee arthroplasty that had been published between 1995 and April 2003. Inclusion criteria were a study of primary total knee arthroplasty, more than 100 knees in the study, provision of baseline data and rating of postoperative outcomes with a standardized symptom scale, and an experimental or quasi-experimental study design. The abstracting form included a list of potential prognostic factors, including comorbidities, radiographic evidence of joint destruction, bone loss, integrity of the extensor mechanism, range of motion, alignment, tibiofemoral angle, and ligament integrity, as well as the characteristics of the operating surgeon, such as procedure volume and experience. RESULTS: Sixty-two studies met the criteria and were reviewed. Total knee arthroplasty was found to be associated with substantial functional improvement, with the effect sizes varying with the measure that was used. Physician-derived measures showed effect sizes of 2.35 and 3.91, whereas patient-derived measures showed smaller effect sizes (1.27 and 1.62). Few investigators used multivariate models to identify associations between outcomes and patient characteristics. CONCLUSIONS: Total knee arthroplasty is a generally effective procedure, but the current English-language literature does not support specific recommendations about which patients are most likely to benefit from it.

The quality of care under a managed-care program for dual eligibles.

PURPOSE: Our objective in this study was to compare the quality of care provided under the Minnesota Senior Health Options (MSHO), a special program designed to serve dually eligible older persons, to care provided to controls who received fee-for-service Medicare and Medicaid managed care. DESIGN AND METHODS: Two control groups were used; one was drawn from nonenrollees living in the same area (Control-In) and another from comparable individuals living in another urban area where the program was not available (Control-Out). Cohorts living in the community and in nursing homes were included. Quality measures for both groups included mortality rates, preventable hospital admissions, and preventable emergency room (ER) visits. For the community group, nursing home admission rates were also tracked. For nursing home residents, quality measures included quality indicators derived from the Minimum Data Set. RESULTS: There were no differences in mortality rates for either cohort. MSHO had fewer short-stay nursing home admissions but no difference for stays 90 days or longer. MSHO community and nursing home residents had fewer preventable hospital and ER visits compared to Control-In. There were no major differences in nursing home quality indicator rates. IMPLICATIONS: The cost of changing the model of care for dual eligibles from a mixture of fee-for-service and managed care to a merged managed-care approach cannot be readily justified by the improvements in quality observed.

Estimating the patient care costs of teaching in a teaching hospital.

PURPOSE: Because leaders at medical schools and teaching hospitals need current data to estimate the clinical costs of graduate medical education, the authors developed a new methodology to estimate the hospital costs associated with the presence of teaching physicians for the year 2002. METHOD: A hospital accounting system was used to determine the case mix-adjusted direct variable costs for 41,522 inpatient admissions associated with or without a teaching physician. RESULTS: Prior to adjustment, teaching cases had greater median costs than non-teaching cases. After severity adjustment, teaching cases in aggregate were associated with an additional 4.4% of the total direct variable cost of inpatient admissions, or US 3.6 million dollars. The size of the teaching effect varied by service, ranging from -5.7% for medical services to 13 percent for behavioral services. The effect of teaching on cost centers such as laboratory, pharmacy, and radiology varied by specialty service. Teaching was associated with a negligible 0.7% relative difference in length of stay. CONCLUSION: The incremental effects of teaching on hospital patient care costs are modest. These analyses can be repeated annually to detect changes in teaching costs and to target areas of excessive cost for interventions that improve efficiency. Our results and methods for identifying hospital costs associated with teaching services may prove useful in negotiations between academic health centers and affiliated teaching hospitals.