Impact of breast cancer treatment guidelines on surgeon practice patterns: results of a hospital-based intervention.

BACKGROUND: Despite evidence regarding the effectiveness of post-surgical treatments for early-stage breast cancer, older women are less likely to receive appropriate therapy. We evaluated the impact of surgeon-specific "performance reports" on adherence to treatment guidelines among older women with breast cancer. METHODS: We obtained diagnostic and treatment data from hospital tumor registries supplemented with self-reported adjuvant therapy information on 1099 patients with stage I or II breast cancer diagnosed between November 1, 1992, and January 31, 1997, at 6 Rhode Island hospitals. We compared rates of appropriate treatment receipt before and after distribution of performance reports. Hierarchical analysis was used to account for the nesting of patients within surgeons. Separate analyses of mastectomy and breast-conserving surgery were performed. RESULTS: Age was negatively associated with post-surgical treatment, with patients who had breast-conserving surgery and who were older than 80 years significantly less likely to undergo radiation therapy (adjusted odds ratio = 0.08 [0.04, 0.14]) or appropriate adjuvant therapies (adjusted odds ratio = 0.14 [0.08, 0.22]) or both relative to 70- to 79-year-old patients. This effect did not improve post-intervention. While there was much variability in compliance with guidelines, surgeons' characteristics did not explain this variation. CONCLUSIONS: In Rhode Island, advanced age continues to be associated with less than adequate breast cancer therapy. Providing surgeons with "feedback" on the appropriateness of adjuvant treatment for older patients was insufficient to alter established practices. Using guideline compliance data as standard "quality indicators" of physician practice may be required.

Cerebrospinal fluid lactate dehydrogenase isoenzyme analysis for the diagnosis of central nervous system involvement in hematooncologic patients.

BACKGROUND: Central nervous system (CNS) involvement is common in hematooncologic diseases. The aim of the current study was to determine the diagnostic value of cerebrospinal fluid (CSF) lactate dehydrogenase (LDH) isoenzyme analysis for the diagnosis of CNS involvement in hematooncologic patients. METHODS: The study was comprised of 63 consecutive hematooncologic patients without previous CNS disease who underwent CSF examination as an integral part of their initial staging procedures (44 patients) or for the evaluation of neurologic symptoms (19 patients). Fifteen of these patients had CNS involvement by leukemia or lymphoma. The LDH isoenzyme pattern was established in the CSF of all patients and analyzed by the classification and regression trees (CART) method to construct a decision tree for the prediction of CNS involvement. An additional group of 30 consecutive patients comprised a validation set that was used for cross-validation of the CART-derived decision tree. RESULTS: A decision tree, with a single split at LDH5 >/= 2.8% for the prediction of CNS involvement, was constructed and validated by data from a validation set of patients. The decision tree had a sensitivity of 93% and a negative predictive value of 98%. One patient (1.6%) and 2 patients (6.6%) were misclassified in the derivation and validation sets, respectively. Overall, in the combined derivation and validation patient population, the decision tree misclassified 3.2% of patients, whereas CSF cytologic examination misclassified 4.3% of patients. CONCLUSIONS: Analysis of the LDH isoenzyme pattern in CSF fluid may be helpful in the evaluation of CNS involvement in patients with hematologic malignancies. The combination of CSF cytology and LDH isoenzyme analysis may improve the sensitivity of CSF cytology significantly.

Sonographically thick placenta: a marker for increased perinatal risk--a prospective cross-sectional study.

The aim of this study was to determine placental thickness by ultrasound examination throughout pregnancy and establish the correlation of sonographically thick placenta with perinatal mortality and morbidity. Placental thickness was determined by routine sonographic examination throughout pregnancy in 561 normal singleton pregnancies. Thick placenta was determined as placenta that was above the 90th percentile. Gravidae between 20-22 weeks' gestation (n=193) and 32-34 weeks (n=73) were then divided into two groups according to placental thickness. The study group consisted of 44 gravidae with thick placenta. The control group included 151 gravidae with placental thickness between the 10th and 90th percentile. A comparison of perinatal mortality and morbidity rates as well as the incidence of small and large for gestational age neonates was conducted.A linear increase of placental thickness was found to correlate with gestational age throughout pregnancy. No statistical differences were observed between the two groups with regard to obstetrical variables such as maternal age, parity and gestational age at delivery. No correlation was found between placental thickness and maternal age or parity. The incidence of perinatal mortality was significantly higher among gravidae with thick placentae (6.82% versus 0.66 per cent, P=0.037, 95 per cent confidence interval 1.71-70.29). Birthweight at term was found to be above 4000 g in 20.45 per cent of the thick-placenta group as compared to 5.3 per cent in the control group (P=0.001, 95 per cent CI 2.08-13.85), and birthweight of less than 2500 g was found in 15. 9 per cent of the thick-placenta group as compared to 7.3 per cent in the control group (P=0.03, 95 per cent CI 1.11-8.14). The incidence of fetal anomalies was 9.1 per cent in the thick-placenta group and 3.97 per cent in the control group (not significant). Sonographically thick placenta is associated with increased perinatal risk with increased mortality related to fetal anomalies and higher rates of both small for gestational age and large for gestational age infants at term.

Lactate dehydrogenase isoenzyme analysis for the diagnosis of pleural effusion in haemato-oncological patients.

This study aimed to evaluate the utility of the pleural fluid lactate dehydrogenase (LDH) isoenzyme algorithm for the differential diagnosis of pleural fluid in patients with haematological malignancies. Twenty consecutive haemato-oncological patients with pleural effusion, hospitalized in the Haematology Department during a 2.75-year period, were prospectively and independently evaluated for the cause of effusion by standard methods for the LDH isoenzyme algorithm. The causes of the pleural effusions established during the standard evaluations were compared to the results obtained from the LDH isoenzyme algorithm. Following the standard evaluation, the pleural effusion was attributed to congestive heart failure in one patient, to infection in six, to the underlying malignancy in 12 and to concomitant congestive heart failure and malignancy in one. LDH isoenzyme analysis correctly predicted the cause of pleural effusion in 18 patients (positive predictive value 90%). In haemato-oncological patients, the pleural fluid LDH isoenzyme pattern may be helpful in the differential diagnosis of the most common causes of pleural effusion.

Facility characteristics associated with hospitalization of nursing home residents: results of a national study.

OBJECTIVES: To test the effect of facility characteristics on the probability of hospitalization of nursing home residents, controlling for resident characteristics and the competing risk of death. RESEARCH DESIGN: Study data were derived from the evaluation of the implementation of the Resident Assessment Instrument, the Minimum Data Set (MDS) in 1993. The data consisted of 2080 residents in 253 NHs as well as the annual On-Line Survey Certification of Automated Records (OSCAR). MEASURES: Multinomial logistic regression was used to determine the effects of selected resident and facility characteristics on hospitalization or death within 6 months of baseline, adjusting for the complex sampling design (using SUDAAN). RESULTS: By controlling for resident demographics, advance directives, diagnoses, selected clinical signs, and type of payer, we found that homes with special care units, more physicians (above the median 0.08 FTE physicians on staff or contract), and any physician extenders (nurse practitioners or physician assistants) were less likely to hospitalize their residents. Homes in which over 3.6% of the residents received respiratory treatment were more likely to hospitalize their residents. CONCLUSIONS: Findings suggest that Medicare HMOs should consider the capacity of nursing facilities, especially in terms of medical care capacity and clinical resources, to limit hospital admissions.

Organizational and environmental factors associated with nursing home participation in managed care.

OBJECTIVE: To develop and test a model, based on resource dependence theory, that identifies the organizational and environmental characteristics associated with nursing home participation in managed care. DATA SOURCES AND STUDY SETTING: Data for statistical analysis derived from a survey of Directors of Nursing in a sample of nursing homes in eight states (n = 308). These data were merged with data from the On-line Survey Certification and Reporting System, the Medicare Managed Care State/County Data File, and the 1995 Area Resource File. STUDY DESIGN: Since the dependent variable is dichotomous, the logistic procedure was used to fit the regression. The analysis was weighted using SUDAAN. FINDINGS: Participation in a provider network, higher proportions of resident care covered by Medicare, providing IV therapy, greater availability of RNs and physical therapists, and Medicare HMO market penetration are associated with a greater likelihood of having a managed care contract. CONCLUSION: As more Medicare recipients enroll in HMOs, nursing home involvement in managed care is likely to increase. Interorganizational linkages enhance the likelihood of managed care participation. Nursing homes interested in managed care should consider upgrading staffing and providing at least some subacute services.

[Quality assurance in the mental health department of the Israel Defense Forces].

A review of quality assurance in the mental health department of the Israel Defence Forces allowed the examination of certain unique elements of quality control which pertain to the military. These include the psychiatric medical board, the computerized documentation of appointments and sessions with soldiers, the psychiatric hospitalization database, control systems implemented in the draft boards, peer-review boards and supervision, and a special officer in charge of handling outside consultations and queries. There were other components of quality assurance and control as well. These instruments are vital in a dynamic system constantly striving to improve clinical performance. Future plans include the continued use and expansion of quality control boards, the inclusion of quality assurance in the curriculum of mental health officers, and the use of clinical guidelines in working with soldiers. All of this is in keeping with the principle of continuous quality improvement, with the aim of viewing the soldier in need of help as a client.

Postacute care following stroke or hip fracture: single services and combinations used by Medicare beneficiaries (1987-1992).

OBJECTIVE: To describe the use of postacute services alone or in combination following a hospitalization for a hip fracture or stroke by Medicare beneficiaries who were relatively well and living in the community prior to the index event. METHODS: Health-service use histories were constructed using Medicare claims. Patients in the study represented all subjects from a 1% sample of Medicare beneficiaries who were age 70 years or older at the time of the index hospitalization. RESULTS: From 1987 to 1992, the proportion of patients receiving any postacute care and those receiving combinations of care increased. For example, there was a doubling of the proportion of patients with either condition using sequences of rehabilitation with home health or SNF and home health. Within 1 year of the hospitalization, 42.6% of patients with stroke and 35.0% post-hip fracture had been rehospitalized. DISCUSSION: Resource use and assessment of patient outcomes should be examined across the continuum of postacute care and in the months beyond to examine the relative effectiveness of different combinations of care.

[Symptom severity among chronic schizophrenics in hospital and in the community].

We compared severity of symptoms of chronic schizophrenics in a psychiatric hospital with those treated in its outpatient clinics. The Positive and Negative Syndrome Scale and the Mini-Mental State examination were used to assess the schizophrenic symptoms and cognitive performance, respectively, of 25 chronic schizophrenic inpatients matched for gender, age and education with 25 chronic schizophrenic outpatients. The Global Assessment Scale and the Clinical Global Impression Scale were used to test global functioning. Assessment included psychiatric and medical history and treatment and demographic characteristics. In-patients had significantly more positive, negative and general psychiatric symptoms. Their cognitive and general functioning were impaired. Most in-patients also had medical problems. Age of onset of schizophrenia among the in-patients was younger. Results show a marked difference in severity of symptoms and level of functioning between chronic schizophrenic in-patients and out-patients. These differences should be considered in the planning of discharge of chronic in-patients from psychiatric hospitals into the community.

Benefits of home health care after inpatient rehabilitation for hip fracture: health service use by Medicare beneficiaries, 1987-1992.

OBJECTIVE: To examine the added benefit of home health services for elderly patients with hip fracture discharged home after inpatient rehabilitation. DATA: Medicare claims from 1% of 1986 beneficiaries followed until 1992. STUDY POPULATION: Persons hospitalized with hip fracture at 70 years or older who had no major Medicare claims during the year before hospitalization and who were discharged home after inpatient rehabilitation. OUTCOMES: Rehospitalization and any nonskilled nursing facility (non-SNF) nursing home admission during the 12 months after hospital discharge. RESULTS: Patients who received additional home health services (27.2%) were less likely to be hospitalized than those who received rehabilitation only (31.1%); they were also less likely to have a non-SNF nursing home admission (11.3% vs 23.3%), and more likely to survive the year with no subsequent Medicare claims (65.6% vs 55%). Propensity scores were used to adjust for nonrandom treatment selection in a Cox proportional hazards analysis showing that home health was associated with a significantly lower risk of nursing home admission (adjusted odds ratio = .42, 95% confidence interval .21-.84), and hospitalization (adjusted odds ratio = .65, 95% confidence interval .26-1.00). CONCLUSIONS: Studies of the relative effectiveness of post-acute services and postdischarge evaluations of inpatient rehabilitation should consider additional home care as a postacute service and examine optimal postacute treatment to minimize additional service use.

Changes in hospitalization associated with introducing the Resident Assessment Instrument.

OBJECTIVE: To compare the rates of hospitalization among cohorts of nursing home residents assembled before and after the implementation of the federally mandated Resident Assessment Instrument (RAI). SAMPLE: Subjects were nursing home residents chosen from 268 facilities in major Metropolitan Statistics Areas in 10 states and representing more than 1500 facilities and 60,000 residents. Two resident cohorts (1990 and 1993) were sampled (8 to 16 residents per facility, depending upon facility size) as part of an evaluation of the impact of implementing the RAI. METHODS: Research nurses reviewed records, interviewed staff, observed patients, and completed an RAI at baseline and 6 months later. All transitions during this interval (hospital admissions, nursing home transfers, returns home, death, etc.) were tracked. Using polytomous logistic regression, we tested the effect of cohort on the probability of being hospitalized in light of the competing risks of dying or remaining in the home, controlling for demographic and casemix variables, and having a DNR order in the chart. RESULTS: A total of 4196 residents were studied, 2118 in 1990 (age 81.3, female 77.7%, LOS 6+ months 49.8%) and 2078 in 1993 (age 81.7, females 75.5%, LOS 6+ months 50.2%). The unadjusted probability of hospitalization dropped from .205 to .151. Multivariate analyses revealed a significant adjusted odds of hospitalization of .74 (95% CI .60-.91) and no cohort effect on home discharge or death. Among severely cognitively impaired residents, the adjusted odds of hospitalization in 1993 compared with the 1990 cohort was 0.74 (.53-1.03). Finally, among survivors in both cohorts who had a follow-up MDS performed, and whose ADL remained stable, 15.9% were hospitalized in 1990, whereas only 10.9% were hospitalized in 1993. On the other hand, ADL decliners were more likely to have been hospitalized in 1993 than in 1990 (40.6% vs 25.2%). CONCLUSIONS: Although other changes in the industry, clinical practice, and health care policy may have influenced hospitalization of nursing home residents, the substantial reductions observed among the cognitively impaired and those with stable ADL suggest superior and uniform assessment information in the form of the RAI contributed significantly to this decline.

Differential diagnosis of pleural effusion by lactate dehydrogenase isoenzyme analysis.

STUDY OBJECTIVE: To determine the diagnostic value of pleural fluid lactate dehydrogenase (LDH) isoenzyme analysis in the differential diagnosis of pleural fluid. PATIENTS AND METHODS: Eighty-seven consecutive patients with pleural effusion caused by congestive heart failure (33), infection (33), and malignancy (21) comprised a derivation set of patients. Pleural fluid LDH activity and isoenzyme pattern were established in all patients and analyzed by the classification and regression trees (CART) method. An additional group of 20 consecutive patients comprised a validation set that was used for cross-validation of CART-derived decision tree. RESULTS: A decision tree, with a positive predictive value of 83%, was constructed and validated by data from a validation set of patients. CONCLUSIONS: Pleural fluid LDH isoenzyme pattern may be helpful for the differential diagnosis of the most common causes of pleural effusions: congestive heart failure, infections, and malignancy.

Impact of the Medicare Catastrophic Coverage Act on nursing homes.

The Medicare Catastrophic Coverage Act (MCCA) of 1988 altered eligibility and coverage for skilled nursing facility (SNF) care and changed Medicaid eligibility rules for nursing-home residents. Detailed data on the residents of a for-profit nursing-home chain and Medicare claims for a 1 percent sample of beneficiaries were used to examine the impact of the MCCA on nursing homes. The case mix of nursing-home admissions was scrutinized, specifically for length of stay, discharge disposition, rate of hospitalization, and changes in payer source. Findings revealed that, although the proportion of Medicare-financed nursing-home care increased, as did the case-mix severity of residents during the MCCA period, there was no corollary reduction in hospital use by nursing-home residents.

Effect of the Medicare Catastrophic Coverage Act on payer source changes among nursing home residents.

The Medicare Catastrophic Coverage Act (MCCA) of 1989 was designed to expand Medicare's post-acute care benefits, reduce copayments, and raise the asset limit for Medicaid eligibility. This analysis uses a semi-Markov transition model to estimate the effect of the MCCA on changes to Medicare coverage and the spend-down rate among 5,551 new nursing home admissions followed for an average of 2.5 years. We found that Medicare use increased in 1989 and the risk of transiting from Medicare to self-pay decreased compared to 1988. Spend-down from self-pay to Medicaid was 60% more likely in 1990. The MCCA clearly increased access to Medicare coverage of nursing home care among individuals previously paying privately.

Trees and splines in survival analysis.

During the past few years several nonparametric alternatives to the Cox proportional hazards model have appeared in the literature. These methods extend techniques that are well known from regression analysis to the analysis of censored survival data. In this paper we discuss methods based on (partition) trees and (polynomial) splines, analyse two datasets using both Survival Trees and HARE, and compare the strengths and weaknesses of the two methods. One of the strengths of HARE is that its model fitting procedure has an implicit check for proportionality of the underlying hazards model. It also provides an explicit model for the conditional hazards function, which makes it very convenient to obtain graphical summaries. On the other hand, the tree-based methods automatically partition a dataset into groups of cases that are similar in survival history. Results obtained by survival trees and HARE are often complementary. Trees and splines in survival analysis should provide the data analyst with two useful tools when analysing survival data.

The outcomes of patients newly admitted to nursing homes after hip fracture.

OBJECTIVES: The outcomes of elderly, hospitalized patients discharged to nursing homes after hip fracture were examined. METHODS: For 2624 hip fracture patients admitted to any of 43 proprietary nursing homes between 1984 and 1988, admission assessments were examined in relation to 1-month outcomes. RESULTS: Mean patient age was 82 +/- 7 y; 85% of the sample were female. Within 1 month after discharge, 24% had returned home, 12% had been rehospitalized, 3% had died, and 61% remained in the nursing home. Characteristics significantly associated with morality included disorientation, functional dependency, neurologic diagnoses, and use of cardiac medications, antidepressants, or narcotics. Rehospitalization was significantly associated with age, gender, living with someone, being ambulatory, and functional dependency. Returning home was associated with younger age, living with someone, being ambulatory, and having no disorientation, functional dependency, or psychiatric or neurologic diagnoses, nor any pressure sores. CONCLUSIONS: Better-functioning persons and those with social support returned home; physically and cognitively impaired persons and those taking narcotics, cardiac medications, or antidepressants were likely to die; and younger men, those with social support, those with functional dependency, and those who were free of disorientation were more likely to be rehospitalized.

Factors affecting conversion rates to Medicaid among new admissions to nursing homes.

OBJECTIVE: This study examines conversion to Medicaid as a payment source among a cohort of newly admitted nursing home residents. DATA SOURCE: The longitudinal data used came from regular assessments of residents in the National Health Corporation's 43 for-profit nursing homes in Missouri, Kentucky, South Carolina, and Tennessee. This information system tracked all residents who were discharged, providing a comprehensive record that may have spanned multiple admissions. STUDY DESIGN: Using survival analysis methods, Cox regression, and survival trees, we contrasted the effect of state, initial payment source, education, age, and functional status on the rate of spend-down to Medicaid. DATA EXTRACTION METHODS: New-admission cohorts were created by linking an admission record for a newly admitted resident with all subsequent assessments and follow-up records to ascertain the precise dates of any payment source changes and other discharge transitions. PRINCIPAL FINDINGS: For the 1,849 individuals who were admitted as self-payers and who were still in the nursing home at the end of one year, there is a 19 percent probability of converting to Medicaid. All analytic methods revealed that education, age, and state of residence were predictive of spend-down among residents who were admitted as self-payers. CONCLUSIONS: Our results confirm the effect of education as an SES indicator and state as a proxy for Medicaid policy on spend-down. Future research should model the effects and duration of intervening hospitalizations and other transitions on Medicaid spend-down among new admissions.

An exploratory analysis of survival with AIDS using a nonparametric tree-structured approach.

We illustrate an analysis with classification and regression trees applied to survival data. Through this application, we provide a description of the opportunistic diseases and sociodemographic factors that contribute to survival among people with human immunodeficiency virus disease. The analyses are based on 43,795 cases reported to the Centers for Disease Control between January 1, 1984, and December 31, 1987. We used vital status as of December 31, 1989, to estimate mortality rates. We identified Kaposi's sarcoma and opportunistic diseases causing central nervous system damage (cryptococcosis, primary lymphoma of the brain, cytomegalovirus disease, and progressive multifocal leukoencephalopathy) as important predictors of death. In addition, advanced age at diagnosis (50+), race (white/other), and history of illicit drug use were found to be important determinants. Estimates of the cumulative probability of survival for subgroups of individuals defined by the tree structure illustrate the effect of these determinants on mortality. For the purpose of comparison, two proportional hazards models were also fit to the data using factors identified in the tree structure as the determinants of interest. This application illustrates the utility and limitations of both this new technique and proportional hazards models for epidemiologic research.