Elsevier

General Hospital Psychiatry

Volume 22, Issue 1, January–February 2000, Pages 1-10
General Hospital Psychiatry

The use of administrative data to assess quality of care for bipolar disorder in a large staff model HMO

https://doi.org/10.1016/S0163-8343(99)00057-2Get rights and content

Abstract

We examined patterns of care for 1246 adults treated for bipolar disorder in a large health maintenance organization. Computerized pharmacy and visit data were used to identify patients treated for bipolar disorder. Automated medical records for the following year were used to assess continuity and dosing of treatment with mood stabilizers, laboratory monitoring for adverse effects and therapeutic serum levels, and frequency of follow-up visits. Of our 1246 subjects 83% filled a mood stabilizer prescription during the 1-year study period, and doses were within recommended ranges 80% of the time. Over 75% of the patients on mood stabilizers had at least one apparent interruption in medication use. Approximately half of the long-term users of mood stabilizers had at least one 7-month period without a recorded blood level and approximately half had a similar period without monitoring for adverse medication effects. Of the 116 subjects discharged from a psychiatric hospitalization 58% had a visit with a psychiatrist or a psychiatric nurse practitioner within 30 days. Of those discontinuing mood stabilizer treatment 68% made a mental health visit within 90 days. Our findings demonstrate the feasibility of using administrative data systems for population-based quality of care assessment and suggest opportunities for improving the care of bipolar patients.

Introduction

Bipolar affective disorder affects approximately 1% of the general population and is associated with substantial morbidity [1], mortality [2], and use of health services 3, 4, 5. One-year prevalence rates from community-based studies range from 1.2% in the Epidemiological Catchment Area Study (ECA) [6] to 1.3% in the National Comorbidity Study [7]. Studies of treated cases suggest much lower prevalence (0.5%) [8]. In earlier research at the Group Health Cooperative of Puget Sound, a large staff model HMO in western Washington, we found a treated prevalence rate of 0.42% of bipolar disorder [9].

Practice guidelines for the treatment of bipolar disorder [10] have emphasized the importance of ongoing treatment with mood stabilizing medications, regular follow-up visits, and regular laboratory monitoring for blood levels and adverse effects of mood-stabilizing medications. Relatively few data are available concerning how well “real world” treatment conforms to these guideline recommendations, and the limited data available suggest considerable room for improvement in the treatment of bipolar disorder. Using automated pharmacy data, a research team at Kaiser Permanente Northwest Region found that lithium use in this HMO was more often intermittent rather than continuous and that discontinuation of lithium was associated with psychiatric hospitalization [11]. That study, however, did not account for switches from lithium to alternative mood stabilizers as an explanation for apparent treatment discontinuation. Another recent study by Katzelnick [12] also described frequent interruptions in mood stabilizer treatment. More data on patterns of care for patients with bipolar disorder are needed to guide future quality improvement efforts.

In this report, we examine the patterns of care for 1246 patients treated for bipolar disorder at the Group Health Cooperative of Puget Sound. Our study had two objectives:

  • 1.

    To demonstrate the feasibility of using computerized administrative data systems to assess quality of care for bipolar disorder across a large population

  • 2.

    To generate information on the quality of care for bipolar disorder in the ‘real world’ setting of this large staff model HMO.

Section snippets

Study setting

Group Health Cooperative (GHC) of Puget Sound is a staff model HMO that provides comprehensive medical care for approximately 400,000 persons in western Washington State. The HMO provides comprehensive care on a capitated basis with most members receiving coverage through employer subsidized plans. The sociodemographic characteristics of the HMO reflect those of people living in the Seattle metropolitan area except for a slightly higher level of education and fewer individuals at the low and

Study sample

Of the 1246 patients in the bipolar study sample, 66% were women. The mean age was 43 (SD 14) with a range from 18 to 89.

Use of psychotropic medications

Table 2 shows the number of subjects who filled prescriptions for various psychotropic medications during the 1-year study period.

A total of 1036 (83% of the 1246 subjects) filled at least one prescription for a mood stabilizing medication, 57% filled a prescription for lithium, 22% for carbamazepine, and 28% filled at least one prescription for valproate, 21% filled

Use of mood stabilizing medications

Our findings regarding the use of mood stabilizers contain both encouraging and discouraging news. Of subjects in our sample 83% received some treatment with mood stabilizing medications during the study period. Mood stabilizer doses were within recommended ranges approximately 80% of the time. When blood levels were monitored, values were typically within recommended ranges. The majority of patients, however, experienced at least one interruption in mood stabilizer treatment. Percent of time

Implications for quality assessment and improvement

Readers may be tempted to make judgments about quality of care for bipolar disorder in the particular health plan studied, in all group/staff model health plans, or in managed care organizations in general, but such judgments may be premature. Some data on continuity of pharmacotherapy in other health plans are available, but differences in methods may easily account for differences in results. We know of no comparable data on populations of patients treated in other systems of care (e.g.,

Acknowledgements

The authors would like to acknowledge the review and suggestions from Joan Bagnall M.D., M.P.H., Azmi Nabulsi, M.D., M.P.H., Jay Chmiel Ph.D., and Kevin Noble M.S. at Abbott Laboratories, and the review of earlier drafts by Michael Quirk, Ph.D. and Neil Baker, M.D. at the Group Health Cooperative of Puget Sound. This work was supported by a grant from Abbott Laboratories.

References (23)

  • W Coryell et al.

    The enduring psychosocial consequences of mania and depression

    Am J Psychiatry

    (1993)
  • F.K Goodwin et al.

    Manic-Depressive Illness

    (1990)
  • M.M Weissman et al.

    Affective disorders

  • G Simon et al.

    Health care utilization by patients treated for bipolar disorder in an insured population

    Psychiatr Serv

    (1999)
  • Wyatt RJ, Henter I: An economic evaluation of manic-depressive illness—1991. Soc Psychiatry Psychiatr Epidemiol...
  • Regier DA, Narrow WE, Rae DS, Manderscheid RW, Locke BZ, Goodwin FK: The de facto U.S. mental and addictive disorders...
  • R.C Kessler et al.

    Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States

    Arch Gen Psychiatry

    (1994)
  • P Bebbington et al.

    The epidemiology of bipolar affective disorder

    Soc Psychiatry Psychiatr Epidemiol

    (1995)
  • J Unützer et al.

    The treated prevalence of bipolar disorder in a large staff model HMO

    Psychiatric Serv

    (1998)
  • Practice guidelines for the treatment of patients with bipolar disorder

    Am J Psychiatry

    (1994)
  • R.E Johnson et al.

    Lithium use and discontinuation in a health maintenance organization

    Am J Psychiatry

    (1996)
  • Cited by (77)

    • Bipolar disorder in primary care: Medication treatment by co-located psychiatrists versus primary care clinicians supported by psychiatrists

      2022, General Hospital Psychiatry
      Citation Excerpt :

      For example, both arms showed essentially unchanged proportions of participants receiving antidepressant or benzodiazepine treatment between initial and last visit. Limitations include possible unmeasured differences between participants, frequency of concurrent PTSD, free-text entry of medication data (lacking reliable dose data), no assessment of treatment interruptions [8], and no data on participant prior medication treatments. Other factors potentially limiting generalizability include patients declining to participate in the trial, and use of research funding for collaborative care training.

    • Correlates of deliberate self-harm among youth with bipolar disorder

      2022, Journal of Affective Disorders
      Citation Excerpt :

      A new episode of BD was defined as: (1) at least 1 inpatient claim or at least 2 face-to-face outpatient claims on different days for a diagnosis of Bipolar I, Bipolar II, or Bipolar NOS/other bipolar spectrum disorder (identified by the International Classification of Disease, 9 and 10 Revision, Clinical Modification [ICD-CM 9 and 10 codes] codes 296.0–296.1, 296.4–296.7, 296.80, 296.81, 296.89, 301.13, F30-F31, F34.0) and (2) no history of BD during the 180-day period prior to the index diagnosis of BD. Previous studies evaluating the validity of using claims data to identify individuals with psychiatric disorders indicate that individuals with two or more outpatient claims for a given disorder are likely to have the disorder of interest (Lurie et al., 1992; Unützer et al., 2000). The follow-up period was limited to 5 years, since follow-up beyond this time would have an over-representation of youth with an earlier age at index date of the new BD episode.

    • Implementing composite quality metrics for bipolar disorder: Towards a more comprehensive approach to quality measurement

      2010, General Hospital Psychiatry
      Citation Excerpt :

      Quality measures need to assess clinically relevant processes of care over which providers have control so that they can inform quality improvement initiatives [15]. Previous studies regarding quality of care for bipolar and other mental disorders in routine care have only focused on adherence to medication treatment guidelines [16] or relied solely on administrative data, which are relatively easy to ascertain [11,14,17]. One of the limitations of administrative data is the lack of information on patient symptoms, provider decision making and details regarding psychosocial treatments, all of which are necessary for constructing patient-centered measures.

    View all citing articles on Scopus
    View full text