The status of sleep abnormalities as a diagnostic test for major depressive disorder.

BACKGROUND: Psychiatry lags other fields in development of diagnostic tests. METHODS: A literature review and meta-analysis was conducted to ascertain if polysomnographic abnormalities (REM density, REM latency, sleep efficiency, slow wave sleep, stage 1 and stage 2 sleep) warrant additional effort to develop them into a clinical diagnostic test for major depressive disorder (MDD). The 31 publications meeting inclusion criteria were then classified into one of three progressive steps using guidelines for evaluating the clinical usefulness of a diagnostic test. RESULTS: Most of the abnormalities found in MDD patients, when compared to healthy controls, occurred in the expected direction with moderate effect sizes but with substantial publication bias and heterogeneity. Eleven studies compared abnormalities in MDD to other psychiatric disorders (step 2a), and four studies provided data on the sensitivity or specificity of the findings in differentiating among the psychiatric disorders that frequently appear on the same differential diagnostic list as MDD (step 2b). No multicenter trial has been conducted prospectively to test the clinical utility of the diagnostic test (step 3). LIMITATIONS: Only published articles in the English language were used. CONCLUSIONS: Sleep studies for the detection of MDD appear replicable with a moderate effect size. However, additional step 1 studies are needed to define the sensitivity and specificity. The heterogeneity of sleep recording, scoring techniques, and MDD must also be addressed.

A four-step approach for developing diagnostic tests in psychiatry.

A four-step approach for developing diagnostic tests in psychiatry is proposed. Step 1, a biological variable is observed to be deviant from healthy controls in a particular patient population. The demonstration of test retest reliability of the finding using blinding procedures is an essential component of this early step. Step 2, is the demonstration of potential clinical usefulness of the specific finding. The two most important objectives at this step are demonstration of difference between the target patient population and appropriate control groups (these should be groups of patients with diagnoses that commonly appear on the differential diagnostic lists of the target disorder). Estimation of the effect size of the finding could be a reasonable guide to which findings should be considered good candidates for Step 3 studies. During Step 3 the performance characteristics of the test should be established. Specifically, the sensitivity, specificity, positive and negative predictive values of the biological marker should be examined. Step 4 defines the clinical application of the test and helps standardize the technique used in large and multicenter clinical trials. Multicenter trials should pave the road towards standardization of laboratory procedures used to conduct the test, as well as providing data regarding cost effectiveness and impact on both short-term and long-term clinical outcomes.

Psychosocial factors and treatment adherence in paediatric HIV/AIDS.

A social ecological model provides a promising framework for understanding the individual, family, and societal factors contributing to non-adherence to treatment of paediatric HIV. This study explored which factors relevant to this model are associated with caregivers' adherence and child health outcomes. A cross-sectional design was utilized to assess relationships among current individual, familial, extra-familial factors, caregiver adherence, and viral load. Data were collected from 43 caregivers, and viral load data were obtained from the medical records of their HIV+ children. Caregiver drug and alcohol use and HIV+ status were associated with non-adherence and elevated viral load. Negative outcome expectancy was associated with lower adherence but was not significant in the multivariate analyses. Family factors were not significant, but these measures had low reliability in this sample. Extra-familial factors such as dissatisfaction with medical specialty care and more stressful life events were not directly associated with adherence but were related to increased caregiver substance use. Results of this first study to explore multiple predictors of adherence and health outcomes in paediatric HIV require replication with larger samples, but findings suggest caregiver characteristics that place children at risk for disease progression due to poor adherence to treatment.

Measuring adherence to treatment of paediatric HIV/AIDS.

Parent, child, physician report and pill counts were used to measure adherence in paediatric HIV. Relationships to viral load were assessed. Pill counts were considered invalid. Adherence measures did not correlate with one another. Physicians reported lower adherence than parents, but parent and physician report correlated with viral load. The clinical and research utility of the various measures are discussed.

Challenges encountered in measuring outcome for a rural psychiatric residential program.

Evaluation of outcome measures can provide policymakers with valuable information on the effectiveness of psychiatric rehabilitation. Two specific challenges in collecting outcome measures for psychiatric rehabilitation programs are heterogeneity of outcomes and difficulty with follow-up. These two challenges were illustrated in the process of evaluating Rose Hill Center, a rural residential psychiatric rehabilitation program. The original design was to conduct interviews with former residents and family members and verify healthcare utilization. The difficulty of locating people and their reluctance to participate conspired to lower the follow-up rates. The design was modified to improve the follow-up rate but decrease the details of specific outcomes. The results showed a high but biased follow-up rate, with more information obtained for people who graduated from the program. The residents with planned discharge showed excellent outcomes in terms of living situation, working situation, healthcare utilization, and low severity of current symptoms. High compliance with medication spoke to the program's philosophy of including the residents in the decision-making process. The dual challenges of heterogeneity of outcomes and difficulty in follow-up that limit efforts to document the value of psychiatric rehabilitation are discussed.

Retrospective review of selective serotonin reuptake inhibitors and falling in older nursing home residents.

We compared the rate of falling in older nursing home residents who had been prescribed selective serotonin reuptake inhibitors (SSRIs), other classes of antidepressants, and no antidepressants. Data were obtained from pharmacy records, medical records, fall logs, and incidence reports for one nursing home (1995 data). Older adults on SSRIs were more likely to fall than older adults not on antidepressants (p = .003) and were more likely to have an injurious fall (p = .03). The association with falling remained significant even when including potential confounders (p = .007). Older nursing home residents should be treated for depression. However, SSRIs may also carry an increased risk for falling.

Gender differences in problem severity at assessment and treatment retention.

Women in treatment for substance abuse have been reported to have more severe problems at assessment than men but not to differ in treatment retention. To examine gender differences in problems at assessment, 30-day retention, and treatment completion, data from Detroit's publicly funded substance abuse treatment system were used. Women had significantly more severe problems at assessment, lower 30-day retention, and lower treatment completion rates than men. These gender differences in retention remained significant even after controlling for problem severity, primary drug of abuse, and referred treatment setting. There was no evidence of improvements in women's problems at assessment or retention over time during this period. Women presented with more severe problems at assessment and were less likely to stay in treatment for 30 days or to complete treatment than men. Monitoring gender differences in problems at presentation and retention outcomes is recommended to assess local need for interventions.

Novelty seeking as a predictor of treatment retention for heroin dependent cocaine users.

This study examined the relationship between novelty seeking between treatment retention and among heroin dependent cocaine users. Participants were treated with buprenorphine maintenance and contingency management. The Tridimensional Personality Questionnaire's (TPQ) Novelty Seeking scale was administered to 68 participants prior to buprenorphine induction. Demographics, mood and anxiety disorders, antisocial personality disorder, and substance use were also assessed. Variables with significant relationships with overall retention were entered into a logistic regression analysis. In addition, using a survival analysis, all variables with significant relationships with time to drop-out were entered into a multivariate proportional hazards regression with time dependent covariates. Results demonstrated that although high novelty seekers, in comparison to low novelty seekers, were more likely to drop-out by the end of treatment, they had higher retention rates during the early phases of treatment. It is suggested that buprenorphine and contingency management were viewed by participants as novel treatment components and thus facilitated high novelty seekers' success early in treatment. If replicated, results suggest that inclusion of novel treatment components might facilitate retention among this at-risk group.

Outcomes of the Learn, Share & Live breast cancer education program for older urban women.

OBJECTIVES: This study assessed whether the Learn, Share & Live breast cancer education program resulted in favorable, replicable, and sustainable outcomes. METHODS: The program was implemented at index (year 1) and replication (year 2) sites. Baseline interviews (year 1; n = 240) and 2 follow-up telephone interviews (years 2 and 3; n = 337 and 323) were used to assess postintervention changes. RESULTS: From baseline to year 2, mammography adherence and stage of adoption improved at the index site relative to the replication site. Knowledge scores and percentages of respondents reporting that a friend had spoken with them about mammography improved significantly. Improvements were sustained through year 3 (2 years postintervention). In year 3, replication participants showed improvements in regard to knowledge and perceived mammography benefits, and there was a trend toward increased adherence. Site differences in postintervention adherence may have stemmed from respective choices of follow-up activities. CONCLUSIONS: The study outcomes affirm the impact of Learn, Share & Live, indicating a replicated and sustained program effect. Future studies should continue longer follow-up and explore the importance of providing mammography opportunities along with education.

Temporal dissociation between lithium-induced changes in frontal lobe myo-inositol and clinical response in manic-depressive illness.

OBJECTIVE: The most widely accepted hypothesis regarding the mechanism underlying lithium's therapeutic efficacy in manic-depressive illness (bipolar affective disorder) is the inositol depletion hypothesis, which posits that lithium produces a lowering of myo-inositol in critical areas of the brain and the effect is therapeutic. Lithium's effects on in vivo brain myo-inositol levels were investigated longitudinally in 12 adult depressed patients with manic-depressive illness. METHOD: Medication washout (minimum 2 weeks) and lithium administration were conducted in a blinded manner. Regional brain myo-inositol levels were measured by means of quantitative proton magnetic resonance spectroscopy at three time points: at baseline and after acute (5-7 days) and chronic (3-4 weeks) lithium administration. RESULTS: Significant decreases (approximately 30%) in myoinositol levels were observed in the right frontal lobe after short-term administration, and these decreases persisted with chronic treatment. The severity of depression measured by the Hamilton Depression Rating Scale also decreased significantly over the study. CONCLUSIONS: This study demonstrates that lithium administration does reduce myo-inositol levels in the right frontal lobe of patients with manic-depressive illness. However, the acute myo-inositol reduction occurs at a time when the patient's clinical state is clearly unchanged. Thus, the short-term reduction of myo-inositol per se is not associated with therapeutic response and does not support the inositol depletion hypothesis as originally posited. The hypothesis that a short-term lowering of myo inositol results in a cascade of secondary signaling and gene expression changes in the CNS that are ultimately associated with lithium's therapeutic efficacy is under investigation.

A survey of prescribing practices for monoamine oxidase inhibitors.

OBJECTIVE: A survey examined prescribing practices for monoamine oxidase inhibitors (MAOIs) and explored reasons for the widely noted decline in their use. METHODS: A one-page questionnaire was sent in 1997 to 1,129 members of the Michigan Psychiatric Association. A total of 717 responses were received, for a response rate of 64 percent. Only data from the 573 psychiatrists who were currently practicing were used. RESULTS: Twelve percent of the respondents never prescribed MAOIs, 27 percent had not prescribed them for at least three years, and 17 percent had prescribed them from one to three years ago. Thirty percent of the respondents had prescribed an MAOI within the past three months, and 14 percent between three and 12 months ago. The most frequent reasons for not prescribing the drugs were side effects and interactions with other medications (46 percent), preference for other medications (30 percent), and dietary restrictions necessary for patients taking MAOIs (19 percent). Ninety-two percent of respondents believed that MAOIs were useful for atypical depression, 64 percent for major depression, 54 percent for melancholic depression, 56 percent for panic disorder, 44 percent for social phobia, 27 percent for dysthymia, 12 percent for obsessive-compulsive disorder, and 19 percent for posttraumatic stress disorder. However, only 2 percent said they would use MAOIs as their first-line treatment in atypical depression, and only 3 percent would use them a first-line treatment in social phobia. CONCLUSIONS: The results document the commonly held view that practicing psychiatrists believe MAOIs are efficacious but use them infrequently, primarily due to concerns about side effects and drug interactions.

Cognitive impairment and depression predict mortality in medically ill older adults.

BACKGROUND: Depression and cognitive impairment are common in medically ill older adults. Few studies, however, have investigated the roles of both in predicting mortality for medically ill older adults. METHODS: We used a cohort of consecutive patients aged 60 or older admitted to a rehabilitation hospital (N = 667) of whom 455 completed a standardized protocol measuring cognition (Dementia Rating Scale), depression (Geriatric Depression Scale), and disabilities (Functional Independence Measure). Burden of medical illnesses was measured with the Charlson Index. Vital status was assessed one year later. RESULTS: Those subjects who did not complete the screening were more likely to die (24% vs 17%; p = .02) during the one-year follow-up. Of those who completed the screening, male sex (odds ratio [OR] = 1.84), depression (mild OR = 1.64; moderate OR = 2.49), and more severe cognitive impairment (OR = 2.13) predicted mortality independent of age, medical illnesses, or disabilities. No interaction of cognitive impairment and depression was detected. In those subjects cognitively intact, moderate depression (OR = 4.95) and male sex (OR = 3.42) were independent risk factors for dying. In those subjects without depression, male sex (OR = 2.24) and elevated Charlson Index (OR = 1.42) predicted mortality. CONCLUSIONS: Depression and cognitive impairment are independent predictors of one-year mortality in this subgroup of medically ill older adults.

Health status and health care costs for publicly funded patients with schizophrenia started on clozapine.

OBJECTIVE: The study examined the effect of clozapine treatment on the health care costs and health status of people with schizophrenia who are supported by public funds. METHODS: Thirty-three patients with schizophrenia hospitalized in a state facility were interviewed within one week of starting clozapine and six months later. Health status was assessed with four clinical rating scales measuring severity of psychopathology, negative symptoms, depression, and quality of life. Cost and health care utilization data were collected for the six months before and after initiation of clozapine. RESULTS: Only 52 percent of the subjects stayed on clozapine for six months. Subjects who continued on clozapine were more likely to be discharged within six months than those who did not continue. Six months after clozapine was started, health care costs showed a sayings of $11,464 per person, even after adjustment for pretreatment costs, and health status was improved. CONCLUSIONS: For subjects who continued on clozapine for six months, clozapine treatment was associated with reduced days of psychiatric hospital care, reduced overall costs despite increased outpatient treatment and residential costs, and improved health status.

Neighbors for a smoke free north side: evaluation of a community organization approach to promoting smoking cessation among African Americans.

OBJECTIVES: This study evaluated a community organization approach that emphasized involvement of audiences in program planning and implementation in promoting nonsmoking among African American residents of low-income neighborhoods. METHODS: The quasi-experimental design involved a 24-month intervention in 3 low-income, predominantly African American neighborhoods in St. Louis. Intervention neighborhoods were compared with comparable, untreated neighborhoods in Kansas City. RESULTS: The program was successful in engaging audience members in its governance and in instigating numerous and diverse neighborhood activities to promote nonsmoking. The prevalence of smoking declined from 34% to 27% in program neighborhoods but only from 34% to 33% in comparison neighborhoods. This difference was apparent within all demographically defined subsamples, indicating that observed changes were consistent and not attributable to confounding by demographic characteristics. CONCLUSIONS: A community organization approach emphasizing local authority for program decisions and involvement of informal networks may have an appreciable impact on smoking among residents of low-income, African American neighborhoods.

Effectiveness of helicopter versus ground ambulance services for interfacility transport.

BACKGROUND: Helicopters provide rapid interfacility transport, but the effect on patients is largely unknown. METHODS: Patients requested to be transported between facilities by helicopter were followed prospectively to determine survival, disability, health status, and health care utilization. A total of 1,234 patients were transported by the primary aeromedical company; 153 patients were transported by ground and 25 patients were transported by other aeromedical services because of weather or unavailability of aircraft. RESULTS: There were no differences at 30 days for survivors in disability, health status, or health care utilization. Nineteen percent of helicopter-transported patients died compared with 15% of those transported by ground (p=0.21). CONCLUSION: The patients transported by helicopter did not have improved outcomes compared with patients transported by ground. These data argue against a large advantage of helicopters for interfacility transport. A randomized trial is needed to address these issues conclusively.

Appropriate imaging in children with skull fractures and suspicion of abuse.

PURPOSE: To define the role of computed tomography (CT) in children aged 2 years and younger with head trauma, suspected abuse, and normal neurologic findings. MATERIALS AND METHODS: From 1992 through 1995, 87 consecutive children with skull fractures visible at plain radiography were referred to child protective services for evaluation of suspected abuse. Their cases were retrospectively reviewed. RESULTS: Of 67 children with normal neurologic findings, 35 (52%) were not referred for CT. No patient in this group developed delayed findings requiring further evaluation. Of the 32 (48%) who underwent head CT, only six (19%) had evidence of acute intracranial injury, despite the presence of minimal depression and stellate, multiple, and diastatic fractures. Of 20 children with acute neurologic findings, 16 (80%) had positive CT scans, which led to neurosurgical intervention in nine (45%). CONCLUSION: No child with normal neurologic findings had a clinically important abnormality depicted at CT. CT scans did not alter clinical management, clinical outcome, or legal outcome. Thus, routine CT of all patients with skull fractures in this population may be unnecessary.

Development of proliferative diabetic retinopathy in African-Americans and whites with type 1 diabetes.

OBJECTIVE: To investigate the comparable risk of developing proliferative diabetic retinopathy (PDR) in African-Americans and whites with type 1 diabetes. RESEARCH DESIGN AND METHODS: Using a cohort design with the sample drawn from medical records, the sample consisted of 312 people with type 1 diabetes (97 African-Americans, 215 whites) having at least two visits to a Model Demonstration Unit with gradeable fundus photographs (stereo, color, 7 standard fields). Excluded were subjects with preexisting or treated PDR or hemoglobinopathy. Masked grading of the fundus photographs was conducted at the Wisconsin Reading Center. RESULTS: At baseline, African-Americans had poorer glycemic control (mean HbA1 of 11.3 vs. 10.0%, P < 0.0001), higher systolic blood pressure (mean of 117 vs. 110 mmHg, P < 0.001), and were older (mean of 26.8 vs. 19.3 years, P < 0.0001) than the white subjects. African-Americans also tended to have slightly longer duration of diabetes and length of follow-up. In the African-Americans, 17.5% developed PDR, compared with 10.2% in the 215 whites, for an odds ratio (OR) of 1.86 (95% CI 0.93-3.70). When adjusted for baseline glycemic control, retinopathy grade, and length of follow-up, race was not a significant risk factor (OR = 0.73, 95% CI 0.30-1.78). CONCLUSIONS: African-Americans with type 1 diabetes may have a higher rate of developing PDR. The observed racial difference, however, is attributable to the presence of a worse risk factor profile, especially to poorer glycemic control. Efforts should be expanded to improve the care for all individuals with poor glycemic control.

Low cholesterol and violence.

OBJECTIVE: The association between violent behavior and low serum total cholesterol levels was examined in a psychiatric inpatient population with diverse diagnoses. METHODS: The study used a case-control design to compare the cholesterol levels of patients in a long-term psychiatric hospital who had a history of seclusion or restraints (N = 20) and those who did not (N = 20). A low cholesterol level was defined as less than 180 mg/dL. RESULTS: A strong association was found between low cholesterol levels and violent behavior (odds ratio = 15.49), an association that was not due to age, race, sex, or diagnosis. The finding was consistent whether mean levels or dichotomized levels of cholesterol were examined. Physical health, cholesterol-lowering medication, current alcohol use, or unusual diets could not explain the results. However, the raw frequency of episodes of seclusion or restraint as an indicator of the frequency of violent behavior was not associated with cholesterol level. Dichotomizing cholesterol levels at 180 mg/dL yielded high sensitivity (90 percent) for predicting violent behavior but at the cost of low specificity (65 percent). CONCLUSIONS: The results support the hypothesis that an association exists between low cholesterol and violent behavior among psychiatric patients but argue against using cholesterol level as a screening tool for predicting violent behavior.