Connecting Research and Practice: Implementation of Suicide Prevention Strategies in Learning Health Care Systems.

The health care systems affiliated with the Mental Health Research Network strive to be learning health care systems that identify and address evidence gaps of importance to clinicians, patients, and funders. This column describes how research guides clinical care and clinical care guides research in the area of suicide prevention as well as some of the challenges of conducting embedded research.

The Mediating Effect of Sleep Disturbance on the Relationship Between Nonmalignant Chronic Pain and Suicide Death.

IMPORTANCE: Few studies have examined the relationship between nonmalignant chronic pain (NMCP) and suicide death, and even fewer have specifically explored what role sleep disturbance might play in the association between NMCP and suicide death. OBJECTIVE: To assess whether sleep disturbance mediates the relationship between NMCP and suicide death. DESIGN: This case-control study included 2,674 individuals who died by suicide between 2000 and 2013 (cases) and 267,400 matched individuals (controls). SETTING: Eight Mental Health Research Network (MHRN)-affiliated healthcare systems. PARTICIPANTS: All cases and matched controls were health plan members for at least 10 months during the year prior to the index date. MAIN OUTCOMES AND MEASURES: Sociodemographic data and diagnosis codes for NMCP and sleep disorders were extracted from the MHRN's Virtual Data Warehouse. Suicide mortality was identified using International Statistical Classification of Diseases and Related Health Problems (ICD)-10 codes from official government mortality records matched to health system records. RESULTS: After accounting for covariates, there was a significant relationship between NMCP and sleep disturbance; those who were diagnosed with NMCP were more likely to develop subsequent sleep disturbance. Similarly, sleep disturbance was significantly associated with suicide death. Finally, a significant indirect effect of NMCP on suicide death, through sleep disturbance, and a nonsignificant direct effect of NMCP on suicide death provide support for a fully mediated model. CONCLUSIONS AND RELEVANCE: There is a need for clinicians to screen for both sleep disturbance and suicidal ideation in NMCP patients and for health systems to implement more widespread behavioral treatments that address comorbid sleep problems and NMCP.

Pilot study of a preliminary criterion standard for prescription opioid misuse.

Multidisciplinary experts created a behaviorally defined preliminary criterion standard definition of probable prescription opioid misuse (PPOM) that could be rated from material found in administrative, pharmacy, and electronic health record databases. They then derived a scoring system to identify PPOM patients requiring referral to a specialist. Experts next rated cases of misuse and nonmisuse. Rater no. 1 correctly differentiated 37 of 40 cases (92.5%); kappa coefficient was .79 (CI: .57, 1.00). Rater no. 2 correctly identified 39 of 40 cases (97.5%); kappa was .94 (CI: .81, 1.00). Kappa for comparing raters was .73 (CI: .49, .98). This preliminary study demonstrates that multidisciplinary raters can use behaviorally based criteria to identify patients with known PPOM from health plan databases.

The association between depressive symptoms and acute exacerbations of COPD.

BACKGROUND: Depression is an important comorbidity for patients with chronic obstructive pulmonary disease (COPD). The association between depression and acute exacerbations of COPD is unknown. This study was designed to determine the frequency of COPD exacerbations in outpatients with and without depressive symptoms. METHODS: In this retrospective cohort study, patients with a primary diagnosis of COPD were followed for 1 year after discharge from a pulmonary rehabilitation program and the frequency of exacerbations was recorded. Upon completion of the program, all patients were administered the Short-Form 36 Health Survey (SF-36), which contains a mental health domain. Patients were classified as having depressive symptoms based on their domain score, which was separately validated in a second population of patients. RESULTS: Of the 194 patients with COPD who completed the pulmonary rehabilitation program, 32 (16.5%) had depressive symptoms. There were no differences in terms of age, race, pack-years, forced expiratory volume in 1 second (FEV(1)), 6-minute walk distance, body mass index, use of supplemental oxygen, use of inhaled steroids, or the Charlson Comorbidity Index between patients with and without depressive symptoms. Patients with depressive symptoms had more exacerbations in the following year (1.91 vs. 1.36; p = 0.02), were 2.8 times more likely to have ever had an exacerbation (95% confidence interval (CI), 1.1-7.3; p = 0.03), and suffered a first exacerbation earlier (148 days compared with 266 days; p = 0.04) than nondepressed patients. CONCLUSIONS: COPD patients with depressive symptoms have a significantly higher risk for exacerbations. Early screening for depression in patients with COPD may help identify those patients at higher risk for subsequent exacerbations.

Impact of bipolar disorder: results from a screening study.

OBJECTIVE: To characterize the extent and impact of bipolar I and II disorders and rapid cycling in a managed care population using both coded diagnostic claims and clinical screening. METHOD: The Mood Disorder Questionnaire (MDQ) was used to identify bipolar disorder among patients attending the psychiatry service of a large Midwestern health system. Suicidal ideation screening questions were also asked, along with a brief set of relevant history and medication questions. Patients scoring positive on the MDQ or identified as bipolar-positive according to DSM-IV criteria by the screening clinician were administered the Work and Social Adjustment Scale and an Employment questionnaire. Descriptive statistics were used to summarize results. The study was conducted from July 2004 to November 2004. RESULTS: Seventy (6.4%) of 1087 patients had bipolar disorder, 59 of whom completed the entire study. For these patients, the mean time with bipolar disorder was 9.3 (SD 10.2) years. The mean length of the current episode was 10.4 (SD 14.4) weeks, with 22% of patients experiencing a mixed episode, 5% manic-predominant, 12% hypomanic-predominant, and 46% a depressive-predominant episode. Twenty-four percent of patients with bipolar disorder were rapid cycling at the time of their visit; for 5 of these patients, rapid cycling was thought to be related to antidepressant use. Sixty-one percent of patients with bipolar disorder were taking an antidepressant; 69.5% were taking a mood stabilizer. Of these patients with bipolar disorder, 19% were evaluated as high suicidality risk, while 47% were considered moderate risk. Bipolar disorder patients reported problems with employment/employability and social adjustment. About one quarter of these patients ranked problems with family and relationships as marked or severe. Fifty percent of these bipolar disorder patients reported missing at least 1 week of work during the past month; 41% reported fearing the loss of their current job due to their emotional state; and 20% reported being fired/laid off during the past 5 years due to their emotional state. CONCLUSIONS: This research documents some of the clinical features and social and labor-force impact of bipolar disorder in a managed care population and adds several dimensions to data published to date. Fully two thirds of our study subjects with bipolar disorder were found to be at substantial risk of suicide, and bipolar disorder patients in this study reported substantial problems with employment/employability and social functioning.

Bipolar disorder detection, ascertainment, and treatment: primary care physician knowledge, attitudes, and awareness.

OBJECTIVE: To describe knowledge, attitudes, and awareness of bipolar disorder detection, ascertainment, and treatment among primary care physicians in a managed care setting. METHOD: Quota sampling was used to obtain 102 completed surveys assessing knowledge, attitudes, and awareness of bipolar disorder from a pool of 350 primary care physicians in a large, vertically integrated Midwestern health system from June 2004 through August 2004. Descriptive statistics were used to characterize the distribution of the study results at the physician level. RESULTS: Primary care physicians are experiencing challenges in diagnosing and treating bipolar patients, who can be difficult and time-consuming. In answering questions about major depressive episode and manic episode symptoms, at least 15% of respondents assessed most symptoms incorrectly. In analyzing 3 case studies, 9%, 11%, and 28% of respondents, respectively, answered all of the questions correctly. When asked which drugs are U.S. Food and Drug Administration-approved for the maintenance treatment of adults with bipolar I disorder, no survey respondent replied correctly for all drugs listed. Importantly, our survey also indicates that these physicians are very willing to refer bipolar patients to psychiatrists for evaluation and treatment, which may help to ensure optimal care. CONCLUSIONS: Opportunities for improvement exist in diagnosing and treating patients with bipolar disorder in the primary care setting, perhaps aided by guidelines, education, and a collaborative care model with psychiatry.

The clinical history and costs associated with delayed diagnosis of bipolar disorder.

The purpose of this reported study was to determine healthcare utilization and costs associated with delayed diagnosis of bipolar disorder. With use of automated data from a large integrated health system in the Midwest, all patients with newly diagnosed bipolar disorder recorded in any inpatient or outpatient encounter from January 1, 2000 to August 31, 2002 were identified. The date of initial diagnosis was the index date. For each patient in the bipolar cohort, 5 comparison patients were randomly selected from the general population of health system members and matched with the bipolar patients by sex, race, and age (-/+ 5 years). Data on healthcare utilization (inpatient, outpatient, emergency department, pharmacy) were collected with a focus on mental health, from January 1, 1990, through 1 year after the index date. The cohort is 62% female and 64% White. Median time between initial mental health diagnosis and bipolar diagnosis was 21 months, with 33% of subjects receiving a bipolar diagnosis within 6 months of their initial mental health diagnosis; however, for 31% of the remaining bipolar subjects, the time of their initial mental health presentation to bipolar diagnosis was 4 years or more. The number and duration of treatment with antidepressants increased as time to bipolar diagnosis increased. Patients with bipolar disorder had at least twice the number of interactions with the healthcare system before the index date than the non-bipolar comparison group. Mean monthly costs before and after bipolar diagnosis were not strikingly different for patients with bipolar disorder, but costs after bipolar diagnosis increased with increasing time to bipolar diagnosis. Bipolar disorder is a costly illness for which the impact on the healthcare system may vary depending on how quickly it is diagnosed. Delays in diagnosis appear related to additional costs after diagnosis.

The Michigan leadership model: developing a management infrastructure.

University of Michigan Health System underwent a number of reduction strategies in the early 1990s to address the rising costs of healthcare. By 2001, an analysis revealed that these strategies negatively impacted employee satisfaction and patient care. A team of nurse managers was charged with redesigning the current support structure for nurse managers. The team conducted an analysis of the current situation and designed a new model called the Michigan Leadership Model comprising both administrative and leadership support positions.