P6272First plus recurrent CV and hospitalization events in the CArdiovascular and Renal Microvascular outcomE study with LINAgliptin (CARMELINA) in patients with type 2 diabetes and cardiorenal disease

Background/Introduction

CARMELINA was a randomized placebo-controlled clinical trial designed to demonstrate the cardiovascular (CV) safety of linagliptin in patients with type 2 diabetes (T2D) and concomitant cardiorenal disease. Despite a particularly elevated CV risk, only limited long-term evidence from randomized controlled trials for safety and efficacy of glucose lowering medications is available for this population.

Purpose

To characterize the effects of linagliptin on net CV disease and hospitalization burden in this population.

Methods

People with T2D and either i) urine albumin creatinine ratio (UACR) >30 mg/g with concomitant CV disease, or ii) estimated glomerular filtration rate (eGFR) <45 ml/min/1.73m2 regardless of UACR, or eGFR ≥45–75 mL/min/1.73m2 and UACR >200 mg/g, were randomized to linagliptin 5 mg or placebo q.d. in a double-blind fashion in addition to standard of care. We assessed the effects of linagliptin versus placebo on all first plus recurrent CV events and all-cause hospitalizations using a using a negative binomial model to account for within-subject correlation.

Results

A total of 6979 participants were enrolled (mean age 66 years, 63% male, eGFR 54.6 ml/min/1.73m2, median UACR 162 mg/g, 59% with history of ischemic heart disease, 27% with history of heart failure (HF)) and followed for a median of 2.2 years. Adding recurrent events increased the number of events for analysis from 5.3–57.5% across CV/HF outcomes and 112.4% for hospitalizations. In analyses of first plus recurrent events, the event rate ratio (95% CI) with linagliptin versus placebo was 0.98 (0.82, 1.16; p=0.78) for 3-point MACE, 1.03 (0.79, 1.35; p=0.83) for myocardial infarction 1.03 (0.83, 1.29; p=0.77) for myocardial infarction plus revascularization, 0.89 (0.65, 1.22; p=0.48) for stroke, 0.94 (0.70, 1.27; p=0.69) for stroke plus TIA, 0.94 (0.75, 1.20; p=0.63) for hospitalized HF, 0.92 (0.77, 1.11; p=0.40) for the composite of CV death or hospitalized HF, and 0.96 (0.87, 1.06; p=0.40) for all-cause hospitalization (Figure).

Conclusion

Linagliptin showed similar risk of either first or recurrent CV or hospitalization events compared with placebo in patients with T2D and cardiorenal disease. These data support the CV safety of linagliptin and, considering the high volume of recurrent events, underscores the significant CV disease burden experienced by patients with T2D and cardiorenal disease

Acknowledgement/Funding

Boehringer Ingelheim and Eli Lilly

Food impaction: etiology over 35 years and association with eosinophilic esophagitis

With the emergence of eosinophilic esophagitis (EoE) as a common cause of food impaction (FI) and a presumed increase in incidence of EoE in the population, the effect on the incidence of FI has not been well described. The aim of this study is to describe the incidence of FI and endoscopic findings in these patients and the association with EoE. A population-based retrospective chart review of the Rochester Epidemiology Project database was performed to identify all patients within Olmsted County that presented with FI from 1976 to 2012. A review of all endoscopic findings, biopsy results, and demographic data was performed. 497 patients were identified with FI from 1976 to 2012. The overall incidence of FI has changed from 1976 to 2012 (Fig. 1) (P < 0.001). The peak incidence of 17.12 per 100,000 people occurred in the time period 1995 to 2000. Both the incidence of comorbid gastroesophageal reflux disease (GERD) and proton pump inhibitor (PPI) use increased over the time period of the study (P < 0.001 for both). Of these patients, 188 (46.7%) had no abnormalities on their endoscopy. The most common endoscopic finding was stricture in 71 (17.6%) patients followed closely by Schatzki's ring in 68 (16.9%) patients. 139 patients had biopsies performed within 2 years of FI and 50 (36.0%) of those were diagnosed with EoE. We present for the first time the changing incidence of FI over the last 35 years in a population-based setting. We also demonstrate the rise of EoE as an important clinical consideration in patients with FI.

Assessing the incidence trend and characteristics of eosinophilic esophagitis in children in Olmsted County, Minnesota

Previous studies reported increased eosinophilic esophagitis (EoE) incidence in children. It is unclear whether this reported increased EoE incidence is true or due to increased recognition and diagnostic endoscopy among children. A population-based study that evaluated EoE incidence in OC, Minnesota, from 1976 to 2005 concluded that EoE incidence increased significantly over the past three 5-year intervals (from 0.35 [range: 0-0.87] per 100,000 person-years for 1991-1995 to 9.45 [range: 7.13-11.77] per 100,000 person-years for 2001-2005). The aim of this study is to assess the change of incidence and characteristics of EoE in children in the same population between 2005 and 2015 and compare the findings to those reported in the previous study. We retrospectively reviewed the electronic medical records from Olmsted Medical Center and Mayo Clinic between 2005 and 2015, using Rochester Epidemiology Project (REP) resources. All children with EoE diagnosis based on the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) guidelines were included. The incidence and characteristics of children with EoE during the study period were compared to those diagnosed between 1995 and 2005. The incidence of EoE in children adjusted for age and sex was 5.31 per 100,000 population person-years in 1995, 15.2 in 2005, and 19.2 in 2015. Change in annual incidence and seasonal variation were not significant, (P = .48) and (P = .32), respectively. Between 2005 and 2015, 73 children received an EoE diagnosis (boys 49; 67%) compared to 16 children (boys 10; 62.5) between 1995 and 2005. Mean (SD) age at diagnosis was 7.5 (5.2) and 12.8 (4.3) years, respectively. Symptoms differed by age of presentation, with vomiting the most common in children younger than 5 years (41.1% and 43.5%) and dysphagia in those older than 5 years (35.6% and 60.9%). The incidence of EoE was not increased for any specific age-group during the study period (P = .49). This study showed increased incidence of EoE in children in Olmsted County between 2005 and 2015 compared to the incidence between 1995 and 2005 (5.31 per 100,000 population person-years in 1995, 15.2 in 2005, and 19.2 in 2015). However, between 2005 and 2015, the change of incidence was not statically significant, (P = .48) despite the steady increase of EGD performed during the same time frame (64 in 2005 to 144 in 2015). By comparing children diagnosed between 2005 and 2015 to those diagnosed between 1995 and 2005, the mean age at diagnosis was younger in the former group, 7.5 versus 12.8 years. Vomiting replaced dysphagia as the most common clinical presentation. Otherwise, the presenting symptom of EoE in children remained consistent across specific age groups.

Outcomes of oesophageal self-dilation for patients with refractory benign oesophageal strictures

BACKGROUND

Current management of refractory benign oesophageal strictures with endoscopic dilations and stenting leads to resolution of dysphagia in only 30% of patients. Oesophageal self-dilation may be an alternative.

AIM

To evaluate the efficacy and safety of oesophageal self-dilation at a tertiary referral centre.

METHODS

We conducted a retrospective review of patients with refractory benign oesophageal strictures who participated in oesophageal self-dilation at Mayo Clinic (Rochester, MN, USA) between 2003 and 2017. Clinical data including stricture characteristics, Dakkak and Bennett Dysphagia Score, number and dates of endoscopies, and complications were collected. A two-tailed paired Student's t test was used to compare the measures of efficacy, with differences considered significant at a 5% probability level.

RESULTS

We identified 52 patients with refractory strictures treated with self-dilation. The median number of endoscopic interventions was reduced from 9.5 (range 5-30) to 0 (range 0-3) within 12 months before and after self-dilation, respectively (P < 0.0001). A median intervention-free interval of 417 days (IQR 256-756 days) was observed. The mean dysphagia score at baseline was 2.5 (95% CI 2.2-2.8) and 0.33 (95% CI 0.11-0.53) after self-dilation. 23 of 27 (85%) patients who received enteral nutrition prior to self-dilation had their feeding tubes removed.

CONCLUSIONS

Oesophageal self-dilation is an effective way of maintaining oesophageal patency in refractory benign oesophageal strictures, with safety comparable to current standard of care. Prospective studies are needed to further validate the role of self-dilation in treatment of refractory benign oesophageal strictures.

Presence of intraepithelial food antigen in patients with active eosinophilic oesophagitis

BACKGROUND

Although eosinophilic oesophagitis (EoE) is putatively mediated by an abnormal response to food antigen, the oesophagus is considered relatively impermeable to large molecules.

AIM

To assess whether food antigens penetrate the oesophageal mucosa in patients with EoE.

METHODS

Anti-gliadin staining was performed in three groups: active EoE, inactive EoE and EoE patients on a low or gluten free diet. To appraise the specificity of our results, we also performed gliadin staining on six patients without oesophageal disease who were consuming gluten. The groups with EoE on gluten also underwent endoscopic infusion with gluten containing soy sauce and repeat biopsies during the endoscopy. We measured eosinophil density, dilated intercellular spaces (on a 0-4+ scale) and gliadin in oesophageal mucosa by immunofluorescence.

RESULTS

Patients with active EoE had significantly greater epithelial density of anti-gliadin staining when compared to inactive EoE (P < 0.0065) and gluten-free patients (P < 0.0008) at baseline and after soy infusion. Gliadin was not detected in non-EoE control patients. The distribution of gliadin was both cytoplasmic and nuclear. There was good correlation of dilated intercellular spaces grade and total gliadin staining intensity (r = 0.577, P = 0.0077). Acute oesophageal perfusion of a commercial gliadin-rich soy sauce did not lead to an increase in gliadin staining in active or inactive EoE.

CONCLUSION

These findings suggest, although do not prove, that antigen penetration in active eosinophilic oesophagitis might be facilitated by impairment of epithelial integrity.

Oesophageal lichen planus: the efficacy of topical steroid-based therapies

BACKGROUND

Oesophageal lichen planus is an idiopathic inflammatory disorder characterized by significant oesophageal stricturing. Oesophageal lichen planus is a rare, difficult to diagnose, and likely an under recognized disease. As a result, there is no standardized approach to therapy and treatment strategies vary.

AIM

To examine the utility of topical steroid therapy (fluticasone or budesonide) in the management of oesophageal lichen planus.

METHODS

A retrospective chart review was conducted of patients diagnosed with oesophageal lichen planus who underwent baseline and follow up endoscopy pre and post topical steroid therapy between 1995 and 2016 at Mayo Clinic, Rochester MN. Average time between upper GI endoscopy was 3.2 months (0.7-11.7). Swallowed steroid preparations included fluticasone 880 μg twice daily or budesonide 3 mg twice daily. Patients were reviewed for symptomatic response to therapy using the Dakkak-Bennett dysphagia score (0-4, no dysphagia to total aphagia). Pre- and post-endoscopic findings were assessed. Additional baseline demographic, endoscopic, and histologic data were also obtained.

RESULTS

We identified 40 patients who met the inclusion criteria. A significant reduction in median dysphagia score from 1 (0-4) to 0 (0-3) after steroid therapy (P < 0.001) was noted. 62% of patients reported resolution of their dysphagia after receiving topical corticosteroids. 72.5% had an endoscopic response to steroid therapy.

CONCLUSION

Topical swallowed budesonide or fluticasone appear to effective treatment for oesophageal lichen planus.

An allergic phenotype and the use of steroid inhalers predict eosinophilic oesophagitis in patients with asthma

BACKGROUND

Patients with eosinophilic oesophagitis (EoE) commonly have asthma and atopy.

AIMS

To determine the predictive factors of EoE in patients with asthma.

METHODS

A retrospective analysis of a large database identified 156 asthma patients with EoE and 276 patients without EoE between 2000 and 2010. Clinical and laboratory characteristics were first analyzed in half of each group. Significant differences and modelling were then applied to the other half of each group in a split half analysis.

RESULTS

Odds ratios and P-values found to predict the presence of EoE in asthma patients were: allergic vs. non-allergic asthma (4.07, <0.01), food allergies (45, <0.01), allergic rhinitis (2.13, =0.01) and peripheral eosinophilia (4.51, <0.01). The use of inhaled corticosteroids was negatively associated with EoE (0.41, <0.01) for asthma patients and (0.37, P < 0.01) for allergic type asthma patients. EoE patients were also younger (27.4 vs. 41.6 years old, P < 0.01). By logistic regression analysis, allergic asthma, presence of peripheral eosinophilia and use of inhaled steroids remained significant. From these parameters, a 3.5-point scoring system model for EoE in asthma was formed with an ROC = 0.787 on split analysis.

CONCLUSIONS

In descending order, peripheral eosinophilia, allergic asthma and allergic rhinitis are associated with EoE in patients with asthma. Steroid inhalers appear to have a protective effect against EoE. An accurate and simple scoring system can be used as a screening tool to predict the presence of EoE in patients with asthma and dysphagia. EoE should be viewed as part of a generalised allergic phenotype rather than isolated oesophageal disease.

Results of ambulatory pH monitoring do not reliably predict response to therapy in patients with eosinophilic oesophagitis

BACKGROUND

The consensus statements for eosinophilic oesophagitis recommend that ambulatory pH monitoring is one means of determining if gastro-oesophageal reflux is the cause of oesophageal eosinophilia and should guide pharmacological therapy.

AIM

To evaluate prospectively the accuracy of pH monitoring as a predictor of endoscopic, histological and symptomatic response in patients with oesophageal eosinophilia.

METHODS

We conducted a prospective trial in which patients with oesophageal eosinophilic infiltration with ≥15 eos/hpf underwent a 24-h pH study and were placed in one of two treatment arms for 6 weeks based on positive or negative results. Patients with abnormal acid exposure were treated with esomeprazole 40 mg twice daily and others were treated with oral viscous budesonide 1 g twice daily. Response to treatment was assessed by oesophageal histology (<5 eos/hpf) and symptoms.

RESULTS

A total of 51 patients were enrolled in the study. The average patient age was 39 years and 31 patients (61%) were male. The average number of eosinophils per hpf, prior to study enrolment was 41.2 (range 15-140, s.d. 27.7). Nineteen (37%) had positive pH studies and 32 (63%) had negative pH studies. Eighteen patients completed treatment with esomeprazole. Only eleven (61%) had histological response and, of these eleven, five (46%) had symptomatic improvement. A total of 28 patients with normal acid exposure completed treatment with budesonide. Only 16 (57%) had histological and 11 (69%) had symptomatic improvement.

CONCLUSION

In this prospective trial of pH-guided treatment, neither positive nor negative results of initial pH monitoring accurately predicted response to therapy.

Impact of endoscopic ultrasonography and physician specialty on the management of patients with esophagus cancer

While endoscopic ultrasonography (EUS) and EUS-guided fine-needle aspiration (EUS-FNA) are the most accurate techniques for locoregional staging of esophageal cancer, little evidence exists that these innovations impact on clinical care. The objective on this study was to determine the frequency with which EUS and EUS-FNA alter the management of patients with localized esophageal cancer, and assess practice variation among specialists at a tertiary care center. Three gastroenterologists, three medical oncologists, three radiation oncologists and four thoracic surgeons were asked to independently report their management recommendations as the anonymized staging information of 50 prospectively enrolled patients from another study were sequentially disclosed on-line. Compared to initial management recommendations, that were based upon history, physical examination, upper endoscopy and CT scan results, EUS prompted a change in management 24% (95% CI: 12-36%) of the time; usually to a more resource-intensive approach (71%), for example from recommending palliation to recommending neoadjuvant chemoradiation therapy. EUS-FNA plus cytology results altered management an additional 8% (95% CI: 6-15%) of the time. Agreement between specialists ranged from fair (intraclass correlation [ICC=0.32) to substantial (ICC=0.65); improving with additional information. Among specialists, agreement was greatest for patients with stage I disease. EUS and EUS-FNA changed patient management the most for patients with stages IIA, IIB or III disease. EUS, with or without FNA, significantly impacts the management of patients with localized esophageal cancer. With respect to the optimal treatment for each patient, agreement among physicians incrementally increases with endoscopic ultrasound results. Specialty training appears to influence therapeutic decision-making behavior.

Making minimally invasive THR safe: conclusions from biomechanical simulation and analysis

The use of smaller surgical incisions has become popularized for total hip arthroplasty (THR) because of the potential benefits of shorter recovery and improved cosmetic appearance. However, an increased incidence of serious complications has been reported. To minimize the risks of minimally invasive approaches to THR, we have developed an experimental approach which enables us to evaluate risk factors in these procedures through cadaveric simulations performed within the laboratory. During cadaveric hip replacement procedures performed via posterior and antero-lateral mini-incisions, pressures developed between the wound edges and the retractors were approximately double those recorded during conventional hip replacement using Charnley retractors (p < 0.01). In MIS procedures performed via the dual-incision approach, lack of direct visualisation of the proximal femur led to misalignment of broaches and implants with increased risk of cortical fracture during canal preparation and implant insertion. Cadaveric simulation of surgical procedures allows surgeons to measure variables affecting the technical success of surgery and to master new procedures without placing patients at risk.

Validation of the Mayo Dysphagia Questionnaire

While multiple instruments characterize upper gastrointestinal symptoms, a validated instrument devoted to the measurement of a spectrum of esophageal dysphagia attributes is not available. Therefore, we constructed and validated the Mayo Dysphagia Questionnaire (MDQ). The 27 items of the MDQ underwent content validity, feasibility, concurrent validity, reproducibility, internal consistency, and construct validity testing. To assess content validity, five esophageal subspecialty gastroenterologists reviewed the items to ensure inclusion of pertinent domains. Feasibility testing was done with eight outpatients who refined problematic items. To assess concurrent validity, 70 patient responses on the MDQ were compared to responses gathered in a structured patient-physician interview. A separate group of 70 outpatients completed the MDQ twice to assess the reproducibility of each item. A total of 148 patients participated in the validation process (78 [53%] men; mean age 62). On average, the MDQ took 6 minutes to complete. A single item (odynophagia) tested poorly with a kappa value of <0.4. Otherwise, the majority of concurrent validity kappa values were in the good to excellent range with a mean of 0.63 (95% CI 0.22-0.89). The majority of reproducibility kappa values were also in the good to excellent range with a median kappa value of 0.76 (interquartile range: 0.67-0.81). Cronbach's alpha values were excellent in the range of 0.86-0.88. Spearman rank correlation coefficients to assess construct validity were also excellent in the range of 0.87-0.98. Thus, the MDQ is a concise instrument that demonstrates overall excellent concurrent validity, reproducibility, internal consistency, and construct validity for the features of esophageal dysphagia.

Selective contracting in managed care: the case of substance abuse treatment

The authors address two critical questions concerning managed care and outpatient substance abuse treatment organizations. Specifically, they consider (1) to what extent selective contracting occurs between managed care firms and treatment providers and (2) what attributes of treatment providers and their operating environments are associated with selective contracting. Using data from a nationally representative sample of outpatient treatment organizations, the authors find evidence of systematic selection. Several indicators of providers' quality and costs, including accreditation status, private ownership, size, and prior experience with managed care, are positively associated with managed care contracting. By contrast, units providing methadone treatment are less likely to be involved in managed care. To a lesser extent, characteristics of treatment providers' operating environment, including extent of competition based on costs and attributes of the Medicaid managed care program, are also positively associated with managed care contracting.

Change in the population of health systems: from 1985 to 1998

This article compares the predictions in Stephen Shortell's 1988 seminal article, The Evolution of Hospital Systems: Unfulfilled Promises and Self-Fulfilling Prophesies, with current data on health systems over a 14-year period from 1985 to 1998. Specifically, we review five of Shortell's predictions related to the horizontal growth of health systems and compare these predictions with empirical data on structural changes in the population of health systems. Our analyses suggest that Shortell's predictions corresponded to much of the actual behavior demonstrated in the population over the past one-and-a-half decades. Support was found for the following: (1) health systems form in two recurring stages; (2) previously unaffiliated hospitals are affiliating with existing systems rather than participating in the creation of new systems; and (3) health systems have evolved into five different strata, each of which represents different shares of the population; such population patterns have important implications for individual hospitals and health systems. By attending to patterns of change in the industry's social structure, hospitals and health systems can determine whether it is likely to continue along past trajectories or whether it shows signs of change that may pave way for the breakdown of existing organizational forms, entry of new organizational players, and the emergence of new governance structures.

Implementing evidence-based medicine: the role of market pressures, compensation incentives, and culture in physician organizations

OBJECTIVES

To assess the extent to which market pressures, compensation incentives, and physician medical group culture are associated with the use of evidence-based medicine practices in physician organizations.

METHODS

Cross-sectional exploratory study of 56 medical groups affiliated with 15 integrated health systems from across the United States, involving 1,797 physician respondents. Larger medical groups and multispecialty groups were overrepresented compared with the United States as a whole. Data are from two sources: (1) surveys of physicians assessing the culture of the medical groups in which they work, and (2) surveys of medical directors and other managerial key informants pertaining to care management practices, compensation methods, and the management and governance of the medical groups. Physician-level data were aggregated to the group level to attain measures of group culture and then merged with the data regarding care management, incentives, and management and governance. Stepwise multiple regression was used to examine the study hypotheses.

RESULTS

As hypothesized, the number of different types of compensation incentives used (cost containment, productivity, quality) was positively associated with the comprehensiveness of care management practices. The degree of salary control (ie, market-based salary grades and ranges versus the use of bookings or fees and individual negotiation) was also positively associated with the deployment of care management practices. As hypothesized, market pressures in the form of percentages of health maintenance and preferred provider organization patients seen were generally positively associated with the use of care management practices. Organizational culture had no association except that a patient-centered culture in combination with a greater number of different types of compensation incentives used was positively associated with greater use of care management practices.

CONCLUSIONS

Both compensation incentives and managed care market pressures were significantly associated with the use of evidence-based care management practices. The lack of association for culture may be due to the relatively amorphous nature of most physician organizations at this point.

Physician-system relationships: stumbling blocks and promising practices

OBJECTIVES

To identify the barriers, facilitators, and potential better practices to achieving physician-system alignment.

METHODS

Interviews using a semi-structured, open-ended protocol were conducted during a total of 18 site visits, each usually 2 days in length, covering multiple topics of physician group-system alignment. Interviews were conducted with members of the target physician group, key leaders of the health care system, and representatives of physicians not in the target group. The summary of the interviews for each of the site visits was analyzed to determine barriers, facilitators, and better practices for achieving more effective relationships between physician groups and health care systems.

RESULTS

A number of barriers to more effective relationships between physician groups and health systems were identified. Barriers related to environment, culture, and information systems were most prevalent. Other major general areas of barriers encountered were physician leadership, group-system relationship, compensation and productivity, care management practices, group strategy, and accountability. Examples of practices that may help to resolve some of these issues were also identified.

CONCLUSIONS

Physician-system relationships can and do cause problems for improving health care. The evidence from the conducted site visits suggests that specific strategies may help improve these relationships but more research is needed in order assess the actual impact of these strategies.

Physician-system alignment: introductory overview

The papers in this Special Supplement are based on research funded by the participating members of the joint Center for Health Management Research (CHMR) and Center for Organized Delivery Systems (CODS), and supported by the National Science Foundation under its Industry-University Cooperative Research Center Program. This 3-year research initiative from 1996 through 1999 involved 69 physician organizations (primarily organized medical groups as opposed to IPAs) associated with 14 organized delivery systems. The groups ranged in size from three to 958 with an average size of 76.4 and a median size of 25.0. Comparisons of the study groups with United States physician groups overall are shown in Table 1. The study groups are larger and more likely to be multispecialty than all groups in the United States. The organized delivery systems range in size from one hospital to 80 hospitals with an average of 21 hospitals per system and a median of 11 hospitals per system. They average 4.6 affiliated medical groups with a range from one to 23. The organized delivery systems range in total revenues in 1998 from $340 million to $6.2 billion with an average of $2.1 billion. All the study systems are not-for-profit. Most are located in single market areas, but several are located in multiple markets. For the most part, they represent some of the larger most experienced organized delivery systems in the country. Among the primary objectives of the study was to identify the factors most strongly associated with physician alignment with the health care system and the consequences for the implementation of evidence-based care management practices. The study was also designed to identify the barriers and facilitators to achieving such alignment and its consequences.

Factors associated with physician involvement in care management

BACKGROUND

Enthusiasm for the concept of care management (CM) has led to unprecedented growth in the number of guidelines and protocols, but provider organizations have struggled to enlist the active support and participation of physicians in CM activities.

OBJECTIVES

To empirically examine the factors influencing physician participation in and attitudes toward CM activities.

METHODS

Data on 1,514 physicians were used to predict physician attitudes toward CM and their perceptions of group CM behaviors. Dependent variables were modeled using two-stage Heckman selection bias models with fixed effects corrections. Independent predictors included physician- and group-level controls as well as six potential CM participation and attitude facilitators.

RESULTS

Physician participation in the implementation phase of CM activities was positively related to participation and attitude. However, physician participation in the development phase may be negatively related to later participation in CM activities. Management involvement in development phase has mixed effects (positive or no effect), but their involvement in the implementation phase was somewhat negatively related to CM participation and attitude. Financial incentives for participation in CM activities and presence of a useful management information system also appeared to be positively related to attitude and participation.

CONCLUSIONS

Appropriate physician and management involvement, as well as financial incentives and useful management information systems may facilitate physician participation in CM activities. Physician involvement in implementation of CM practices appears to be important, whereas their involvement in the development phase may be negatively related to later attitudes and participation. The findings call for a more in-depth understanding of the timing of physician input in CM activities.

The ties that bind: interorganizational linkages and physician-system alignment

OBJECTIVES

To examine the association between the degree of alignment between physicians and health care systems, and interorganizational linkages between physician groups and health care systems.

METHODS

The study used a cross sectional, comparative analysis using a sample of 1,279 physicians practicing in loosely affiliated arrangements and 1,781 physicians in 61 groups closely affiliated with 14 vertically integrated health systems. Measures of physician alignment were based on multiitem scales validated in previous studies and derived from surveys sent to individual physicians. Measures of interorganizational linkages were specified at the institutional, administrative, and technical core levels of the physician group and were developed from surveys sent to the administrator of each of the 61 physician groups in the sample. Two stage Heckman models with fixed effects adjustments in the second stage were used to correct for sample selection and clustering respectively.

RESULTS

After accounting for sample selection, fixed effects, and group and individual controls, physicians in groups with more valued practice service linkages display consistently higher alignment with systems than physicians in groups that have fewer such linkages. Results also suggest that centralized administrative control lowers physician-system alignment for selected measures of alignment. Governance interlocks exhibited only weak associations with alignment.

CONCLUSIONS

Our findings suggest that alignment generally follows resource exchanges that promote value-added contributions to physicians and physician groups while preserving control and authority within the group.

Physician commitment to organized delivery systems

BACKGROUND

Health care systems have developed many types of contracting vehicles with physicians. The immediate aim of these vehicles has been to foster physician commitment and alignment to the system. The ultimate aim of these vehicles has been to garner managed care contracts, reduce costs, and improve quality. To date, most of these vehicles have failed to improve physician commitment. This may be one reason why the ultimate outcomes have not been observed. Consequently, systems are experimenting with new methods to partner with physicians. One new method is to segment physicians into tightly linked and loosely linked strategic alliances and devote different levels of resources and attention to each.

OBJECTIVES

This study evaluates whether the segmentation of physicians into tightly linked versus loosely linked strategic alliances improves the commitment of physicians to the system. The study then investigates which constituent elements of the tightly linked strategic alliances exhibit the greatest association with commitment. DESIGNS AND SUBJECTS: The study uses a cross-sectional design and survey data drawn from 1,965 physicians affiliated with 14 health care systems around the country. Tightly linked physicians typically practiced in hospital-sponsored group practices, whereas loosely linked physicians typically used the system's hospitals as their primary site of inpatient practice.

MEASURES

Commitment is measured by seven different scales drawn from the literature on organizational commitment, loyalty, and identification. Some of the scales refer to physician attitudes, whereas others describe physician behaviors. The literature suggests that commitment is associated with both instrumental/utilitarian considerations (eg, older age, tenure with system, admissions to system, receipt of a stipend, etc.) as well as administrative involvement/participation considerations (eg, decision-making roles). A series of physician background and practice characteristics are used here to model these two types of factors.

RESULTS

The study finds small but significant differences in commitment between physicians in tightly linked versus loosely linked alliances. Multivariate analyses suggest that instrumental/utilitarian factors (eg, age, receipt of stipend, percent of admissions to the system) may exhibit stronger associations with commitment than the physician's administrative involvement in the organization.

CONCLUSIONS

To the degree that physician commitment is possible, systems should appeal to physicians' calculative motivations using extrinsic rewards rather than normative involvement in the organization.

Risk assumption and physician alignment with health care organizations

OBJECTIVES

To examine the association between risk assumption by individual physicians and physician groups and the degree of alignment between physicians and health care systems.

METHODS

A cross sectional comparative analysis using a sample of 1,279 physicians practicing in loosely affiliated arrangements and 1,781 physicians in 61 groups closely affiliated with 14 vertically integrated health systems. Measures of physician alignment were based on multiitem scales validated in previous studies and derived from surveys sent to individual physicians. Measures of risk assumption were developed from surveys sent to the administrator of each of the 61 physician groups in the sample and to physicians affiliated with these groups. Two stage Heckman models with fixed effects adjustments in the second stage were used to correct for sample selection and clustering respectively.

RESULTS

After accounting for selection, fixed effects, and group and individual controls, physicians in groups with larger proportional revenue from managed care displayed greater normative commitment and system loyalty than physicians in groups with lower proportional managed care revenue. Individual-level managed care risk was also positively related to both normative commitment and group behavioral commitment to the system. Physicians in groups with larger physician equity positions expressed lower levels of normative commitment to the system. Physician productivity compensation was negatively related to all measures of alignment. Finally, group emphasis on individually-based incentives for staff physicians was negatively related to system identification.

CONCLUSIONS

Our findings suggest that organizations must balance individually-based risk schemes with those that emphasize the performance of the group and the system to achieve long-term goals of loyalty, identification, and commitment to the system.

CEO perceptions of competition--and strategic response in hospital markets

Physician-organization integration (POI) has emerged as a key issue for hospitals and health systems seeking to improve the quality and cost-effectiveness of care. Although competition and managed care are often cited as primary market drivers of the adoption of POI strategies, prior research has shown only weak associations between these market attributes and POI. This article argues that the role of key organizational decision makers has not been adequately accounted for in explaining strategic change. The study examines the role of hospital CEO perceptions of competition in predicting the adoption of five different approaches to POI. CEO perceptions of general market competition are explained by a combination of market and organizational attributes. Furthermore, when controlling for objective characteristics of the environment and organization, CEO perceptions of competition have consistent, statistically significant associations with four of five measures of POI examined.

Changes in the structure, composition, and activity of hospital governing boards, 1989-1997: evidence from two national surveys

Hospital governance arrangements affect institutional policymaking and strategic decisions and can vary by such organizational attributes as ownership type/control, size, and system membership. A comparison of two national surveys shows how hospital governing boards changed in response to organizational and environmental pressures between 1989 and 1997. The magnitude and direction of changes in (1) board structure, composition, and selection; (2) CEO-board relations; and (3) board activity, evaluation, and compensation are examined for the population of hospitals and for different categories of hospitals. The findings suggest that hospital boards are engaging in selective rather than wholesale change to meet the simultaneous demands of a competitive market and traditional institutional orientations to community, the disenfranchised, and philanthropic service. Results also suggest parallel increases in collaboration between boards and CEOs and in board scrutiny of CEOs.

Just how integrated are integrated delivery systems? Results from a national survey

This article examines three emergent processes in physician-hospital integrated delivery systems (IDSs). We find these processes are underdeveloped based on data gathered from a national sample of hospitals drawn from nine health care systems. These processes are also loosely coupled with the structures used to integrate physicians and hospitals, as well as with the environmental context in which they occur. Such loose coupling entails both advantages and disadvantages for IDSs.

Case management and quality of life: assessing treatment and outcomes for clients with chronic and persistent mental illness

OBJECTIVE

To examine the impact of treatment setting and exposure to case management services on the quality of life of U. S. veterans with chronic and persistent mental illness.

DATA SOURCES/STUDY SETTING

Data were collected longitudinally on a panel of 895 clients enrolled in 14 pilot programs in Department of Veterans Affairs long-term psychiatric hospitals by the Serious Mental Illness Treatment Research and Evaluation Center during the period 1991-96.

STUDY DESIGN

Data were collected using two primary survey instruments (clinician assessment and client assessment) at baseline, every six months for the first two years, and every year thereafter, for a total of four years of follow-up. Case management exposure over time and its impact on the client's quality of life represent the key variables in the study. Additional controls included a variety of sociodemographic, socioeconomic, and psychiatric characteristics.

DATA COLLECTION/EXTRACTION METHODS

Hierarchical linear modeling was used to control for potential selection bias, test for the compositional effect of treatment setting, and examine the impact of case management exposure over time on the individual client's quality of life.

PRINCIPAL FINDINGS

Increased exposure to case management results in an improved quality of life across several domains, including both objective and subjective dimensions for health, general, leisure, and social, and the subjective dimension only for housing.

CONCLUSIONS

The study findings provide managers, clinicians, and policymakers a fuller understanding of how this mode of service delivery-case management-affects several domains of quality of life for clients with chronic illnesses.

Strategies for effective management participation in community health partnerships

This article develops guidelines for effective health services management participation in community health partnerships. Drawing on our study of Community Care Network (CCN) Demonstration, the strategic alliance literature, and other research, we describe six challenges that health services managers are likely to face as partnership participants and discuss the strategies that they might use to deal with them.

Managed care and outpatient substance abuse treatment intensity

This study examines the extent to which managed care behavioral controls are associated with treatment intensity in outpatient substance abuse treatment facilities. Data are from the 1995 National Drug Abuse Treatment System Survey, a nationally representative survey that includes over 600 provider organizations with a response rate of 86%. Treatment intensity is measured in three ways: (1) the number of months clients spend in outpatient drug treatment, (2) the number of individual treatment sessions clients receive over the course of treatment, and (3) the number of group treatment sessions clients receive over the course of treatment. After accounting for selection bias and controlling for market, organization, and client characteristics, there is no significant relationship between the scope of managed care oversight and treatment intensity. However, the stringency of managed care oversight activities is negatively associated with the number of individual and group treatment sessions received over the course of treatment.

The effects of structure, strategy and market conditions on the operating practices of physician-organization arrangements

Research to date has documented weak or inconsistent associations between market and organizational factors and the adoption of physician-organization arrangements (POAs) (e.g. physician-hospital organizations, management service organizations and independent practice associations) designed to increase physician integration. We argue that POAs may mask considerable variation in how these entities are operated and governed. Further, because the operating policies and practices of POAs are likely to influence more directly the behaviour of physicians than the structural form of the POA, they may be more sensitive to the market and organizational contingencies that encourage integration. This study attempts to test empirically the relative effects of POA type and market, strategic and organization factors on the operating policies and practices of market-based POAs. Results suggest that type of POA, and market, strategic and organizational factors affect risk sharing, physician selection practices, physician monitoring practices and ways in which monitoring information is used to influence physician behaviour in POAs.

Hospital provision of institutional long-term care: pattern and correlates

PURPOSE

This study examined the pattern and correlates of institutional long-term care provision among U.S. community hospitals, differentiating two categories of services: (1) skilled nursing and rehabilitation (SN-R) and (2) other long-term care (O-LTC).

DESIGN AND METHODS

Multinomial logistic regression analysis was used to examine the associations of hospital and community characteristics with the pattern of long-term care provision (SN-R only, O-LTC only, both SN-R and O-LTC, and None) among 3,842 hospitals.

RESULTS

The pattern of long-term care provision was significantly associated with hospitals' mission (for-profit and teaching status) and their internal and external resources.

IMPLICATIONS

Results suggest the importance of considering hospital and community characteristics in predicting the impact of policy changes and in envisioning the role of hospitals in long-term care.

Prevalence, pathophysiology, and treatment of patients with asthma and gastroesophageal reflux disease

About one third of the US adult population experiences symptoms of gastroesophageal reflux on a monthly basis. Asthma is present in about 5% of the same population. This article reviews and summarizes the literature in the following areas: (1) prevalence of gastroesophageal reflux disease (GERD) in asthmatic patients based on clinical symptoms, endoscopic esophagitis, and 24-hour ambulatory esophageal pH recordings; (2) proposed pathophysiologic mechanisms linking the 2 diseases; and (3) medical and surgical treatment trial results of antireflux therapy for asthmatic patients. Asthmatic patients appear to have an increased prevalence of GERD symptoms and 24-hour esophageal acid exposure. The clinical management of these patients remains controversial. Common management approaches to GERD in asthmatic patients include medical therapy with a proton pump inhibitor and/or antireflux surgery, which improve asthma symptoms in many patients but minimally affect pulmonary function.

Community accountability among hospitals affiliated with health care systems

The shift from local, community-based hospitals to more complex, multilevel delivery systems raises questions about the community accountability exercised by hospitals. A national sample of community hospitals is the basis of this study, which examines the ways that community accountability is exercised by the governing boards of hospitals affiliated with health care systems and how such institutions compare with hospitals not affiliated with a health care system. Results indicate that hospitals display community accountability in a variety of ways. Boards of system-affiliated hospitals exercise community accountability most strongly in their information monitoring and reporting activities, whereas free-standing hospitals exercise community accountability through the structural and compositional attributes of their boards. Further, hospitals affiliated with different types of systems vary in the style and degree of accountability they demonstrate.

Medical and psychosocial services in drug abuse treatment: do stronger linkages promote client utilization?

OBJECTIVE

To examine the extent to which linkage mechanisms (on-site delivery, external arrangements, case management, and transportation assistance) are associated with increased utilization of medical and psychosocial services in outpatient drug abuse treatment units.

DATA SOURCES

Survey of administrative directors and clinical supervisors from a nationally representative sample of 597 outpatient drug abuse treatment units in 1995.

STUDY DESIGN

We generated separate two-stage multivariate generalized linear models to evaluate the correlation of on-site service delivery, formal external arrangements (joint program/venture or contract), referral agreements, case management, and transportation with the percentage of clients reported to have utilized eight services: physical examinations, routine medical care, tuberculosis screening, HIV treatment, mental health care, employment counseling, housing assistance, and financial counseling services.

PRINCIPAL FINDINGS

On-site service delivery and transportation assistance were significantly associated with higher levels of client utilization of ancillary services. Referral agreements and formal external arrangements had no detectable relationship to most service utilization. On-site case management was related to increased clients' use of routine medical care, financial counseling, and housing assistance, but off-site case management was not correlated with utilization of most services.

CONCLUSIONS

On-site service delivery appears to be the most reliable mechanism to link drug abuse treatment clients to ancillary services, while referral agreements and formal external mechanisms offer little detectable advantage over ad hoc referral. On-site case management might facilitate utilization of some services, but transportation seems a more important linkage mechanism overall. These findings imply that initiatives and policies to promote linkage of such clients to medical and psychosocial services should emphasize on-site service delivery, transportation and, for some services, on-site case management.

Endoscopic evaluation of the esophagus in infants and children immediately following intraoperative use of transesophageal echocardiography

OBJECTIVE

Intraoperative transesophageal echocardiography (TEE) has evolved as an essential technique for use during pediatric cardiac surgery; however, few studies have evaluated the safety of TEE in children. This series reports endoscopic examination of the esophagus following intraoperative TEE in pediatric patients.

METHODS

Fifty children undergoing congenital heart surgery underwent flexible esophagoscopy that was performed after completion of their heart surgery and after the removal of the transesophageal echo probe. The patients' ages ranged from 4 days to 10 years old, and their weight ranged from 3.0 to 39.8 kg, with a mean weight of 12.6 kg.

RESULTS

Thirty-two of 50 patients (64%) had abnormal results shown on esophageal examinations; this occurred more frequently in the subset of patients weighing < 9 kg. No long-term feeding or swallowing difficulties were noted in any of the 48 patients who survived.

CONCLUSIONS

Intraoperative TEE in infants and children frequently caused mild mucosal injury. Care must be exercised in the insertion and manipulation of the probes.

Managing hospitals in turbulent times: do organizational changes improve hospital survival?

OBJECTIVE

To examine (1) the degree to which organizational changes affected hospital survival; (2) whether core and peripheral organizational changes affected hospital survival differently; and (3) how simultaneous organizational changes affected hospital survival.

DATA SOURCES

AHA Hospital Surveys, the Area Resource File, and the AHA Hospital Guides, Part B: Multihospital Systems.

STUDY DESIGN

The study employed a longitudinal panel design. We followed changes in all community hospitals in the continental United States from 1981 through 1994. The dependent variable, hospital closure, was examined as a function of multiple changes in a hospital's core and peripheral structures as well as the hospital's organizational and environmental characteristics. Cox regression models were used to test the expectations that core changes increased closure risk while peripheral changes decreased such risk, and that simultaneous core and peripheral changes would lead to higher risk of closure.

PRINCIPAL FINDINGS

Results indicated more peripheral than core changes in community hospitals. Overall, findings contradicted our expectations. Change in specialty, a core change, was beneficial for hospitals, because it reduced closure risk. The two most frequent peripheral changes, downsizing and leadership change, were positively associated with closure. Simultaneous organizational changes displayed a similar pattern: multiple core changes reduced closure risk, while multiple peripheral changes increased the risk. These patterns held regardless of the level of uncertainty in hospital environments.

CONCLUSIONS

Organizational changes are not all beneficial for hospitals, suggesting that hospital leaders should be both cautious and selective in their efforts to turn their hospitals around.

Consequences of organizational change in U.S. hospitals

Organizational change has become commonplace among U.S. hospitals. Empirical investigations of the consequences of organizational change, however, are relatively scarce, and findings of existing studies are inconsistent. In this article, the authors review the rationale and performance implications of hospital organizational change in three areas: (1) the development of new multi-institutional arrangements, (2) change in traditional ownership and management configurations, and (3) diversification in organizational products/services and consolidation of organizational scale. Empirical research on hospital change published between 1980 and 1999 in the health services research, social science, and business literatures is reviewed to highlight the potential pitfalls that hospitals may encounter in their effort to remain viable. The article also summarizes the strengths and weaknesses of current hospital change research and provides specific suggestions for future research in this area.

How other members of the top management team see the nurse executive

This study compared how top-level organizational groups within Veterans Affairs Medical Center (VAMC) top management teams perceived nurse executives' participation in strategic decision making. Nurse executives' perceptions of their own involvement in strategic decision making generally agreed with the views of both chiefs of staff and directors. However, the associate director's perceptions of the nurse executive's involvement often differed significantly with those of the nurse. Nurse executives will be better prepared to function effectively within the top management team if they are aware of patterns in how they are perceived by other members.

Physician and clinical integration among rural hospitals

The pressures for closer alignment between physicians and hospitals in both rural and urban areas are increasing. This study empirically specifies independent dimensions of physician and clinical integration and compares the extent to which such activities are practiced between rural and urban hospitals and among rural hospitals in different organizational and market contexts. Results suggest that both rural and urban hospitals practice physician integration, although each emphasizes different types of strategies. Second, urban hospitals engage in clinical integration with greater frequency than their rural counterparts. Finally, physician integration approaches in rural hospitals are more common among larger rural hospitals, those proximate to urban facilities, those with system affiliations, and those not under public control.

An exploratory analysis of market-based, physician-organization arrangements

The growth of physician-organization arrangements (e.g., PHOs, MSOs) by physician specialty groups and delivery organizations is widely acknowledged. Yet, little is known about these new organizational forms. This paper presents a descriptive analysis of physician-organization arrangements (POAs) using survey data obtained from 79 organizations. The analysis examines the extent to which healthcare delivery organizations are engaged in the development of POAs, the number of physicians involved in POAs, and the decision-making control and strategies associated with POAs. These attributes are examined for the sample as a whole and relative to the institutional and market conditions facing study participants. Results are discussed in terms of their implications for health services managers.

Multihospital system affiliation as a survival strategy for rural hospitals under the prospective payment system

The introduction of Medicare's Prospective Payment System (PPS) has disproportionately increased financial pressures on rural hospitals and posed challenges to the survival of these institutions. Increasingly, rural hospitals are seeking strategies that can enhance their chances for survival in a turbulent and hostile environment. This study examined the survival effects of one such strategy, multihospital system affiliation. Specifically, we assessed: (1) whether and how different types of system affiliation in the post-PPS era affect the likelihood of rural hospital survival; (2) whether particular structural, environmental and hospital performance characteristics moderate the effects of system affiliation on rural hospital survival; and (3) whether systematic selection by rural hospitals into multihospital systems potentially accounts for observed relationships between system affiliation and survival. Proportional hazards analyses indicate that system affiliation with investor-owned systems significantly reduces survival probabilities of rural hospitals. Affiliation with not-for-profit systems or system affiliation under contract management arrangements does not affect survival probabilities of rural hospitals. These general findings are moderated by the effects of hospital ownership and size at the time of affiliation. Finally, study findings indicated that systematic selection by poor performing rural hospitals into investor-owned systems has occurred in the post-PPS era. No evidence of selection into not-for-profit systems was discovered.

State legislation and policy affecting rural hospital conversion and closure

Many rural hospitals have closed or converted to organizations that provide health services other than general, acute inpatient care. However, little is known about why rural hospitals convert rather than close. This study evaluates the relationship between state policy and rates of rural hospital conversion relative to rates of rural hospital closure. The expectation was that the relationship among state policies and rates of conversion and rates of closure differs as a function of whether a state policy facilitates rural hospital transition to alternative models of care or supports them in their existing form. National data were analyzed for all rural hospitals from 1984 to 1991. Results indicate that state policies have done little to facilitate widespread adoption of conversion among rural hospitals. Despite these findings, results also indicate that some state policies have resulted in an increase in the rate of rural hospitals conversions as an alternative to closure.

Diversification behavior of multihospital systems: patterns of change, 1983-1985

Studies of diversification in the health care sector have typically focused on hospital-level activity. Little attention has been given to the broader organizational context in which many hospitals operate (multihospital systems, parent holding companies). This article empirically examines diversification behavior of 100 multihospital systems between 1983 and 1985 in order to assess how systems respond to resource constraints, uncertainty, and competitive pressures on inpatient care. Patterns of diversification are identified, and their implications for systems strategic behavior are discussed.

The effects of treatment team diversity and size on assessments of team functioning

Team-based health care assumes that groups representing multiple disciplines can work together to implement care plans that are comprehensive and integrated. It also assumes that professionals can function effectively in an interdependent relationship with members of other occupational groups. However, we know little about what makes effective team functioning. This article examines the factors related to health care team functioning, with specific emphasis on team demographic composition and size. Hierarchical linear modeling is used to analyze 106 Veterans Affairs (VA) hospitals. Results indicate that individuals who operate on more heterogenous and larger teams have lower perceptions of team functioning.

Financial and organizational determinants of hospital diversification into subacute care

OBJECTIVE

To examine the financial, market, and organizational determinants of hospital diversification into subacute inpatient care by acute care hospitals in order to guide hospital managers in undertaking such diversification efforts.

STUDY SETTING

All nongovernment, general, acute care, community hospitals that were operating during the years 1985 through 1991 (3,986 hospitals in total).

DATA SOURCES

Cross-sectional, time-series data were drawn from the American Hospital Association's (AHA) Annual Survey of Hospitals, the Health Care Financing Administration's (HCFA) Medicare Cost Reports, a latitude and longitude listing for all community hospital addresses, and the Area Resource File (ARF) published in 1992, which provides county level environmental variables.

STUDY DESIGN

The study is longitudinal, enabling the specification of temporal patterns in conversion, causal inferences, and the treatment of right-censoring problems. The unit of analysis is the individual hospital.

KEY FINDINGS

Significant differences were found in the average level of subacute care offered by investor-owned versus tax-exempt hospitals. After controlling for selection bias, financial performance, risk, size, occupancy, and other variables, IO hospitals offered 31.3 percent less subacute care than did NFP hospitals. Financial performance and risk are predictors of IO hospitals' diversification into subacute care, but not of NFP hospitals' activities in this market. Resource availability appears to expedite expansion into subacute care for both types of hospitals.

CONCLUSIONS

Investment criteria and strategy differ between investor-owned and tax-exempt hospitals.

Do market-level hospital and physician resources affect small area variation in hospital use?

This study evaluates the effect of market-level physician and hospital resources on hospital use. It is anticipated that higher hospital discharges are associated with (1) greater hospital and physician resources, (2) more differentiated hospital and physician resources, and (3) higher levels of teaching intensity in the community. Data on 14 modified diagnostically related groups (DRGs) and 58 hospital market communities in Michigan are analyzed during a 7-year period. Findings indicate that physician resources, hospital resources, differentiation of hospital and physician resources, and teaching intensity contribute only modestly to discharges, holding constant the socioeconomic attributes of the community and adjusting for the variation in hospital use over time. With the inclusion of hospital and physician resource variables, socioeconomic factors remain important determinants of the variation across market communities. Findings are discussed in terms of their implications for health care organizations, managed care programs, and cost control efforts in general.

Physician involvement in management and governance: the moderating effects of staff structure and composition

Many hospitals are actively pursuing strategies that integrate physicians into their management and governance structures. Despite expectations that these strategies improve hospital efficiency, empirical studies have failed to provide consistent evidence that physician involvement in hospital management and governance improves hospital efficiency. This article examines factors that may moderate the relationship between physician participation in hospital management and governance and hospital efficiency.

On-site primary care and mental health services in outpatient drug abuse treatment units

Providing health services to drug abuse treatment clients improves their outcomes. Using data from a 1995 national survey of 597 outpatient drug abuse treatment units, this article examines the relationship between these units' organizational features and the degree to which they provided onsite primary care and mental health services. In two-stage models, Joint Commission on Accreditation of Healthcare Organizations-certified and methadone programs delivered more on-site primary care services. Units affiliated with mental health centers provided more on-site mental health services but less direct medical care. Units with more dual-diagnosis clients provided more on-site mental health but fewer on-site HIV/AIDS treatment services. Organizational features appear to influence the degree to which health services are incorporated into drug abuse treatment. Fully integrated care might be an unattainable ideal for many such organizations, but quality improvement across the treatment system might increase the reliability of clients' access to health services.