Donald M. Steinwachs

Development and application of a population-oriented measure of ambulatory care case-mix.

This article describes a new case-mix methodology applicable primarily to the ambulatory care sector. The Ambulatory Care Group (ACG) system provides a conceptually simple, statistically valid, and clinically relevant measure useful in predicting the utilization of ambulatory health services within a particular population group. ACGs are based on a persons demographic characteristics and their pattern of disease over an extended period of time, such as a year. Specifically, the ACG system is driven by a persons age, sex, and ICD-9- CM diagnoses assigned during patient-provider encounters; it does not require any special data beyond those collected routinely by insurance claims systems or encounter forms. The categorization scheme does not depend on the presence of specific diagnoses that may change over time; rather it is based on broad clusters of diagnoses and conditions. The presence or absence of each disease cluster, along with age and sex, are used to classify a person into one of 51 ACG categories. The ACG system has been developed and tested using computerized encounter and claims data from more than 160,000 continuous enrollees at four large HMOs and a states Medicaid program. The ACG system can explain more than 50% of the variance in ambulatory resource use if used retrospectively and more than 20% if applied prospectively. This compares with 6% when age and sex alone are used. In addition to describing ACG development and validation, this article also explores some potential applications of the system for provider payment, quality assurance, utilization review, and health services research, particularly as it relates to capitated settings.

Classifying trauma severity based on hospital discharge diagnoses: validation of an ICD-9CM to AIS-85 conversion table

This report describes the development and validation of a computerized system for converting ICD-9CM rubrics to Abbreviated Injury Scale (AIS) scores. In collaboration with the Committee on Injury Scaling of the Association for the Advancement of Automotive Medicine, AIS-85 scores were assigned to 2,062 injury-related ICD-9CM rubrics. To validate the conversion table, AIS and Injury Severity Scores (ISS), derived using the conversion, were compared with those obtained by reviewing the complete medical record for 1,120 trauma cases. Percent agreement in maximum AIS scores (MAXAIS) ranged from 48% for head/neck injuries to 74% for extremity injuries. In 68% of the cases, grouped ISS scores (one to 12; 13 to 19; 20+) were in agreement. Previous studies of the interrater reliability of AIS coding directly from the medical charts have shown that agreement in MAXAIS scores ranges, on average, between 62% for head/neck injuries to 76% for extremity injuries. Grouped ISS scores agree, on average, 75% of the time. The results show that while the computerized conversion is not perfect, it provides reasonably good information on severity that might otherwise be unavailable for large population-based research and evaluation. This paper discusses the potential applications of the conversion table with specific attention to its use in evaluating the extent of trauma care regionalization.

National Outcomes of Cataract Extraction: Increased Risk of Retinal Complications Associated with Nd.-YAG Laser Capsulotomy

PURPOSE The authors studied 57,103 randomly selected Medicare beneficiaries who underwent extracapsular cataract extraction in 1986 or 1987 to determine the possible association between performance of neodymium (Nd):YAG laser capsulotomy and the risk of subsequent retinal break or detachment. METHODS Cases of cataract surgery were identified from Medicare claims submitted in 1986 and 1987 and were followed through the end of 1988. Episodes of cataract surgery, posterior capsulotomy, and retinal complications were ascertained based on procedure and diagnosis codes listed in physician bills and hospital discharge records. Lifetable and Coxs proportional hazards models were used to analyze the risk of retinal detachment or break in patients undergoing and not undergoing capsulotomy during the period of observation. RESULTS Of the 57,103 persons identified as having undergone extracapsular cataract extraction in 1986 or 1987, 13,709 subsequently underwent Nd:YAG laser capsulotomy between 1986 and 1988. A total of 337 persons had aphakic or pseudophakic retinal detachments between 1986 and 1988 and an additional 194 underwent repair of a retinal break. Proportional hazards modeling shows a 3.9-fold increase in the risk of retinal break or detachment among those who underwent capsulotomy (95% confidence interval: 2.89 to 5.25). Younger patient age, male sex, and white race also were associated with increased risk of retinal complications after extracapsular cataract extraction. CONCLUSION The authors conclude that there is a statistically significant increase in the risk of retinal detachment or break in those patients who undergo capsulotomy after cataract extraction. Therefore, capsulotomy should be deferred until the patients impairment caused by capsular opacification warrants the increased risk of retinal complications associated with performance of capsulotomy.

Hospital at Home: Feasibility and Outcomes of a Program To Provide Hospital-Level Care at Home for Acutely Ill Older Patients

Context Hospital care for older people often means iatrogenic complications and a decline in function. Home hospital care might reduce these adverse outcomes. Content Patients were 65 years of age or older and required hospital care for pneumonia, heart failure, chronic obstructive pulmonary disease, or cellulitis. In phase I, they were hospitalized. In phase II, they could choose home hospital care (continuous nursing care followed by at least daily visits from a nurse and a physician). Sixty percent of patients chose home hospital care. Patients who received this type of care had shorter stays; fewer procedures, consultations, and indwelling devices; less delirium; greater satisfaction; and similar functional outcomes. Cautions The study was nonrandomized, and data were missing. Conclusion Home hospital care may be a good alternative for selected patients. The Editors Although the acute care hospital is the standard venue for providing acute medical care, it is expensive and may be hazardous for older persons, who commonly experience functional decline, iatrogenic illness, and other adverse events during care (1-3). Providing acute hospital-level care in a patients home is an alternative to hospital care (4, 5). Although several hospital-at-home models have been studied, there is controversy regarding the effectiveness of this method. In part, this reflects heterogeneity among hospital-at-home models (6). A recent Cochrane review examined surgical and medical early hospital discharge models, terminal care, and admission avoidance, that is, substitutive models. Overall, no differences were found in health outcomes. Patients, but not caregivers, had increased satisfaction with hospital-at-home care, and there was some evidence that substitutive models may be cost-effective (7). However, with some exceptions (8), most of these models would be difficult to distinguish from augmented skilled nursing services, community-based long-term care, or home-based primary care services in the United States. In addition, most studies have been done in countries with single-payer national health insurance systems (7-14). Previous research in the United States has been limited to a pilot study of a physician-led substitutive hospital-at-home model for older persons with acute medical illness (15). The aim of our study was to evaluate the safety, efficacy, clinical and functional outcomes, patient and caregiver satisfaction, and costs of providing acute hospital-level care in a hospital at home that substituted entirely for admission to an acute care hospital for older persons. Methods Patients The target sample was community-dwelling persons, age 65 years and older, who lived in a catchment area and who, in the opinion of a physician not involved in the study, required admission to an acute care hospital for 1 of 4 target illnesses: community-acquired pneumonia, exacerbation of chronic heart failure, exacerbation of chronic obstructive pulmonary disease, or cellulitis. Patients were required to meet validated criteria of medical eligibility for hospital-at-home care (16), which were designed to identify patients who would be medically suitable for this type of treatment. The most common reasons for medical ineligibility were uncorrectable hypoxemia (oxygen saturation <90%), suspected myocardial ischemia, and presence of an acute illness, other than the target illness, for which the patient was required to be hospitalized. Study Design This study was a prospective quasi-experiment conducted in 2 consecutive 11-month phases. During the acute care hospital observation phase (1 November 1990 to 30 September 2001), eligible patients were identified and followed through usual hospital care. Study coordinators verified the patients eligibility for hospital-at-home care using a standard protocol at the time of enrollment. During this observation phase of the study, most patients were identified the morning after admission. These patients made up the acute hospital observation comparison group. During the intervention phase (1 November 2001 to 30 September 2002), eligible patients were identified at the time of admission and were offered the option of receiving their care in hospital at home rather than in the acute care hospital. Patients who chose hospital-at-home treatment were never admitted to the acute care hospital but received treatment, after initial evaluation (usually in the emergency department), in their home. The intervention group comprised all patients eligible for hospital-at-home care, irrespective of where they were treated. Approval The institutional review boards from each study site, the coordinating center, and officials at the Center for Health Plans and Providers at the Centers for Medicare & Medicaid Services (CMS) gave their approval for the study. All participants provided informed written consent. Study Sites The study was conducted in 3 Medicare managed care (Medicare + Choice) plans at 2 sites and at a Veterans Administration medical center. Univera Health and Independent Health, in Buffalo, New York, are Medicare + Choice plans that operate in an independent practice association model. These 2 plans collaborated to provide hospital-at-home care and made up 1 study site (site 1). The Fallon Health Care System (site 2), in Worcester, Massachusetts, operates a not-for-profit Medicare + Choice plan, and the Fallon Clinic, a for-profit multispecialty physician group, provides care on a capitated basis to Medicare + Choice beneficiaries. The Portland, Oregon, Veterans Administration Medical Center (site 3) is a quaternary care and teaching facility. Assessments Age, gender, and primary diagnosis were obtained for all eligible patients. Informed written consent was required for all additional data collection: medical record review, cost data review, and interviews. Staff trained at the coordinating center used standard procedures outlined in a detailed training manual to conduct all interviews, assessments, and medical record reviews. At baseline, interrater reliability was verified among the staff. Quality checks of medical record reviews were done midway through the study. Interrater reliability for ratings on the components of the daily patient interview was confirmed in 13 paired observations ( = 0.91). Interrater reliability among study sites was similar. Medical Record Reviews Medical records were abstracted by using a standardized instrument that captured illness acuity, health status, medication use, results of laboratory tests, treatments, the hospital course and complications, health outcomes, and whether treatment standards were met. Illness acuity was determined by using the Acute Physiology and Chronic Health Evaluation II (APACHE II) score (17). Health status was measured by using clinical indicators appropriate to the diagnoses, a checklist of chronic medical conditions, and the Charlson comorbidity index (18). Medication use was defined as the number of prescribed medications taken on a daily basis at the time of admission. Diagnostic and therapeutic interventions were categorized as potentially difficult or not difficult to do in the home. The patients clinical course was characterized according to whether emergency situations (those that required physician evaluation within 30 minutes, such as the development of acute shortness of breath) or critical complications (death, transfer to intensive care setting, intubation, or myocardial infarction) occurred. Clinical care was examined with regard to completion of illness-specific standards of care (19-21). Eligible patients who consented to participate completed a baseline interview that included demographic information, self-reported health status, assessment of sleep, Katz activities (22), Lawton instrumental activities of daily living (23), continence, mobility, the Geriatric Depression Scale (24), Jaeger vision test, the Mini-Mental State Examination (MMSE) (25), the Digit Span Test (26), and evaluation by the Confusion Assessment Method (CAM) (27). Subsequently, patients were evaluated daily until discharge by using a structured interview consisting of the MMSE, Digit Span Test, and CAM rating. A family member, caregiver, or person who knew the patient well was interviewed at the time of admission to complete the modified Blessed Dementia Rating Scale (28). At 2 weeks after admission, patients and family members were interviewed by telephone to obtain the patients current functional status and to assess his or her satisfaction with care. Intervention: The Hospital-at-Home Model of Care The hospital-at-home model of care has been described previously (15). Briefly, a patient requiring admission to the acute care hospital for a target illness was identified in an emergency department or ambulatory site and his or her eligibility status was determined. Nonstudy medical personnel, usually emergency department physicians, made the decision to hospitalize the patient. All patients who were offered but who declined hospital-at-home care were admitted to the acute care hospital. After informed consent was obtained, the patient was transported home by an ambulance. Patients were evaluated by the hospital-at-home physician either in the emergency department or shortly after arriving at home. Patients who required oxygen therapy were sent home with a portable oxygen apparatus pending delivery of home oxygen therapy. The hospital-at-home nurse met the ambulance at the patients home. The patient had subsequent direct one-on-one nursing supervision for an initial period of at least 8 hours at site 3 and for a period of 24 hours at sites 1 and 2. When direct nursing supervision was no longer required, the patient had intermittent nursing visits at least daily. The hospital-at-home physician made at least daily home visits and was available 24 hours a day for urgent or emergent visits. Nursing and other care components, such as durable medical equipme

Mental health service utilization by African Americans and whites: The Baltimore Epidemiologic Catchment Area Follow-Up

OBJECTIVE To compare mental health service utilization and its associated factors between African Americans and whites in the 1980s and 1990s. DESIGN Household-based longitudinal study with baseline interviews in 1981 and follow-up interviews from 1993 to 1996. SETTING The Baltimore Epidemiologic Catchment Area (ECA) Follow-Up. SUBJECTS Subjects included 1,662 adults (590 African Americans and 1,072 whites). MAIN OUTCOME VARIABLE Use of mental health services, defined as talking to any health professional about emotional or nervous problems or alcohol or drug-related problems within the 6 months preceding each interview. RESULTS In 1981, crude rates of mental health service use in general medical (GM) settings and specialty mental health settings were similar for African Americans and whites (11.7%). However, after adjustment for predisposing, need, and enabling factors, individuals receiving mental health services were less likely to be African American. Mental health service use increased by 6.5% over follow-up, and African Americans were no longer less likely to report receiving any mental health services in the 1990s. African Americans were more likely than whites to report discussing mental health problems in GM settings without having seen a mental health specialist. They were less likely than whites to report use of specialty mental health services, but this finding was not statistically significant, possibly because of low rates of specialty mental health use by both race groups. Psychiatric distress was the strongest predictor of mental health service use. Attitudes positively associated with use of mental health services were more prevalent among African Americans than whites. CONCLUSIONS Mental health service use increased in the past decade, with the greatest increase among African Americans in GM settings. Although it is possible that the racial disparity in use of specialty mental health services remains, the GM setting may offer a safety net for some mental health concerns of African Americans.

National Outcomes of Cataract Extraction: Retinal Detachment and Endophthalmids after Outpatient Cataract Surgery

Background: A near-total shift to cataract extraction on an outpatient basis occurred as a result of an administrative ruling by the Health Care Financing Administration. No national study has been conducted to assess the possible effects of that decision on clinical outcomes of surgery. The authors compared the rates of retinal detachment (RD) repair and hospitalization for endophthalmitis after extracapsular cataract extraction (ECCE) (including phacoemulsification) in 1986 and 1987 with those following inpatient cataract extraction in 1984. Methods: Using the 5% random sample of Medicare beneficiaries, we analyzed the claims of all individuals 66 years of age or older who underwent ECCE by nuclear expression or phacoemulsification in 1986 and 1987. A total of 57,103 patients were identified and followed to the end of 1988. Cumulative probability of RD repair and hospitalization for endophthalmitis was calculated by standard lifetable methods. These findings were compared with the cumulative probability of the same complications in a cohort of 330,000 patients who underwent cataract extraction on an inpatient basis in 1984. Results: In the 1986-to-1987 cohort, the cumulative probability of RD within 3 years after cataract surgery was 0.81% and the cumulative probability of endophthalmitis within 1 year was 0.08%. The rate of RD is similar to that which we previously reported for 330,000 patients who underwent inpatient surgery in 1984, but the rate of endophthalmitis is significantly lower in the 1986-to-1987 outpatient cohort (0.08% versus 0.12%; z = 2.42; P = 0.01). Conclusions: The shift to outpatient cataract surgery was accompanied by no significant increase in the probability of RD repair and possibly a significant decrease in the rate of hospitalization for endophthalmitis.

Measuring need for mental health services in a general population.

This article presents measures of need for mental health services estimated from the 1981 Eastern Baltimore Mental Health Survey, one of five sites participating in the NIMH Epidemiologic Catchment Area Program. Data were collected on the prevalence of specific psychiatric disorders, as determined by the standardized Diagnostic Interview Schedule (DIS), functional status, personal characteristics, patterns of medical and mental health care, and sources of care used. Need is based on mental health services use in the prior 6 months or the presence of two or more manifestations of emotional problems: a) one or more DIS disorders present in the past 6 months, b) a General Health Questionnaire (GHQ) score of four or more current symptoms, or c) the respondents report of having been unable to carry out usual activities in the past 3 months for at least 1 entire day because of an emotional problem. Approximately 14% of adults met the criteria for need, half of whom had made no mental health visits in the prior 6 months and were considered to have unmet need. Need for care was influenced by a variety of sociodemographic and economic characteristics: it was low among the aged and high among persons living alone and the poor on Medicaid. The proportion of need that was unmet varied less but was relatively large for two groups, the aged and nonwhites. Those on Medicaid through public assistance were more likely to have their need met than the near poor.

Variation in Cataract Surgery Practice and Clinical Outcomes

PURPOSE To examine associations between surgical technique, patient and surgeon characteristics, and clinical outcomes of cataract surgery. METHODS Seventy-five ophthalmologists were recruited from three cities based on a sampling scheme stratified by surgeon-reported annual volume of cataract surgery. Seven hundred seventy-two patients undergoing first eye cataract surgery were enrolled, with complete preoperative, perioperative, and 4-month postoperative clinical data on 717 patients (93%). RESULTS Sixty-five percent of surgery was performed by phacoemulsification and 35% by standard extracapsular (ECCE) techniques. Performance of ECCE was associated with the presence of ocular comorbidity and 21 or more years in practice of the surgeon. Performance of phacoemulsification was associated with annual volume of cataract surgery, wherein high-volume (201-399 patients annually) and very high-volume (> 400 patients annually) surgeons had 3.7 and 3.9 times the likelihood of performing phacoemulsification compared with moderate-volume (51-200 cases annually) surgeons. The rates of intraoperative, perioperative, and 4-month postoperative adverse events and the amount of improvement in visual acuity did not differ either by surgical technique or volume stratum. The reported occurrence of posterior capsular opacification within 4 months of surgery was increased in the presence of cortical opacification, one city, and patients operated on by either high- or very high-volume surgeons. CONCLUSIONS In this cohort, no difference in clinical outcomes, as measured by change in visual acuity or occurrence of postoperative adverse events (except for posterior capsular opacification), can be attributed to performance of phacoemulsification versus ECCE or to the reported annual volume of cataract surgery of the surgeon. Self-reported high and very high annual volume of cataract surgery is associated independently with performance of phacoemulsification and surgeons report of posterior capsular opacification at 4 months after cataract surgery.

Detecting and treating retinopathy in patients with type I diabetes mellitus. A health policy model

Diabetic retinopathy is the major cause of new cases of blindness among working-age Americans. The authors analyzed the medical and economic implications of alternative screening strategies for detecting retinopathy in a diabetic population. The approaches compared included dilated fundus examination at 6-, 12-, and 24-month intervals with and without fundus photography. Potential savings from screening and treatment are based on amounts paid by the federal government for blindness-related disability. Screening for and treating retinopathy in patients with type I diabetes mellitus was cost-effective using all screening strategies. Between 71,474 and 85,315 person years of sight and 76,886 and 94,705 person years of reading vision can be saved for each annual cohort of patients with type I diabetes mellitus when proper laser photocoagulation is administered. This results in a cost savings of

Measuring provider continuity in ambulatory care: an assessment of alternative approaches.

62.1 to