Skip to main content

19-12-2011 | Mental health | Article

Double discharge HDR algorithm identifies DSM-IV-TR bipolar disorder


Free abstract

MedWire News: Study findings suggest that identification of DSM-IV-TR bipolar disorder for use in epidemiologic research is possible using a hospital discharge register (HDR)-based algorithm employing two discharge diagnoses of bipolar disorder.

The authors explain that HDRs are often used in epidemiologic research, but uncertainty remains as to the validity of several important psychiatric diagnosticentities, including bipolar disorder.

"While the National Quality Assurance Register for Bipolar Disorder (BipoläR) registry should be regarded as the gold standard for a bipolar disorder diagnosis in Swedish registers with very high specific and detailed phenotypic information, usage is still limited by a low ascertainment rate," say Mikael Landen (Institute of Neuroscience and Physiology, Molndal, Sweden) and colleagues.

They add: "By contrast, the proposed algorithm for identifying bipolar cases in the HDR has complete nationwide coverage and therefore provides greater sample size and statistical power."

For the study, the researchers identified 135 potential cases of bipolar disorder in the Swedish national HDR using two hospital discharge diagnoses of bipolar disorder according toInternational Classification of Diseases (ICD) versions 8-10 during 1973-2004. These search criteria were defined as algorithm A.

False-positive cases were then scrutinized, and the initial algorithm was modified to improve the positive predictive value (PPV), while minimizing false-negatives (algorithm B).

Both algorithms were then externally validated by linking HDR diagnostic data with best-estimate clinical diagnoses from the BipoläR registry, dispensed lithium prescriptions from the National Prescribed Drug Register (PDR), and ICD-10 diagnoses from the National Outpatient Register (NOR).

Of the 135 patients identified as having bipolar disorder using algorithm A, 110 cases were true-positives (positive predictive value=0.81).

Application of algorithm B to the 25 false-positive cases showed that 16 of these cases were attributable to diagnostic codes ICD-8 296.20 (manic-depressive psychosis, depressed type) and/or ICD-9 296.B (unipolar affective psychosis, melancholic form). Algorithm B further improved the PPV obtained with algorithm A to 0.92.

External validation of identified cases according to algorithms A and B and general population controls matched on age and birth year in the BipoläR registry, the PDR, and the NOR, showed that representation was significantly higher using algorithm B compared with algorithm A.

Writing in the journal Acta Psychiatrica Scandinavica, the team concludes: "We suggest that the search algorithm for bipolar disorder caseness based on two separate inpatient episodes according the Swedish HDR is sufficiently sensitive and specific to be used in further epidemiological and genetic research of bipolar disorder, provided that ICD-8 296.20 (manic-depressive psychosis, depressed type) and/or ICD-9 296.B (unipolar affective psychosis, melancholic form) are excluded."

MedWire ( is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2011

By Ingrid Grasmo

Related topics