Concept: Depression - Defining in Administrative Data

Background

This concept describes how MCHP has identified and defined depression in its research. The term depression covers a spectrum of mood disorders that can range from being mild and transitory to a persistent state of incapacitation. One end of the spectrum can be difficult to distinguish from normal reaction and at the other end there is an overlap into severe psychotic disorders. In the classification of mental disorders using the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), major depression is defined as a period of at least two weeks characterized by at least five severe and persistent depressive symptoms. The symptoms must not represent a normal grief reaction or be secondary to an organic cause such as a physical illness or drug exposure. As well as feelings of sadness, symptoms can include; changes in sleep pattern; loss of energy; change in appetite; difficulty concentrating; feelings of worthlessness; and suicidal thoughts. Despite this very precise definition it is not easy to accurately measure the true prevalence of depression. Largely this is due to the considerable variation in clinical presentations but also because of inconsistencies between screening and diagnostic tools as well as in coding methods.

Depression is estimated using hospital discharge abstracts, medical services/physician claims and drug utilization data at MCHP. This concept is based on work done for the 2003/04 Mental Illness Disorders deliverable (Martens et al., 2004) , the Key Findings of the Child Health Atlas 2004 project (Brownell et al., 2004), the Manitoba RHA Indicators Atlas 2009 (Fransoo et al., 2009) , the Profile of Metis Health Status and Healthcare Utilization in Manitoba deliverable (Martens et al., 2010) , and the Health Inequities in Manitoba deliverable (Martens et al., 2010) .

This concept also provides links to relevant SAS code. Please see the SAS code and formats section containing the following links to: SAS code for defining drugs, SAS Code for Depression - Mental Illness Deliverable 2004, and SAS Code for Depression Prevalence - RHA Indicators Atlas 2009 - (internal access only).

IMPORTANT NOTE: In 2012, MCHP research combined the depression and anxiety conditions into "Mood and Anxiety Disorders". The concept titled Mood and Anxiety Disorders - Measuring Prevalence describes the MCHP research approach in this subject area.

Methods

Definition:

Depression, or possibly broader - mood disorders - since we will also capture some people in here that actually are bipolar but whom cannot be separated out in physician / medical services data. Only individuals 10 and older were included in the definition of depression.

1. Martens et al. (2004) and Doupe et al. (2008)

In Martens et al. (2004), depression was defined as the presence of ICD codes from the hospital files or physician claims file for affective psychoses, neurotic depression, adjustment reaction, or depressive disorder, or the ICD code for neurotic disorders plus a prescription for an antidepressant or mood stabilizer. The presence of an electroconvulsive therapy (ECT) procedure code was tested in the definition, but as it did not add any new cases to the numerator, it was removed. Depression was identified through the presence of any hospital discharge abstracts or medical services/physician claims data coding depression using the following definitions. Note that individuals with only a neurotic disorder must also have a prescription:


From the hospital files ( Note: In Martens et al. (2004), this also included the MHMIS files):

• Any of ICD-9-CM codes 296.2-296.8 (affective psychoses), 300.4 (neurotic depression), 309 (adjustment reaction), or 311 (depressive disorder)
• ICD-9-CM code 300 (neurotic disorders) plus a prescription for an antidepressant or mood stabilizer.

From the physician files:

• Any of ICD-9-CM codes 296, 309, or 311
• ICD-9-CM code 300 plus a prescription for an antidepressant or mood stabilizer.

The presence of an electroconvulsive therapy (ECT) procedure code (94.27) was originally included in the definition, but as it did not add any new cases to the numerator, it was removed.

2. Fransoo et al. (2009), Martens et al. (2010) and Chartier et al. (2012)

In the deliverables Manitoba RHA Indicators Atlas 2009 by Fransoo et al. (2009), Profile of Metis Health Status and Healthcare Utilization in Manitoba: A Population-Based Study by (Martens et al., 2010) and Health and Healthcare Utilization of Francophones in Manitoba by Chartier et al., 2012 depression was defined as residents aged 10 or older over a five-year period with any one of the following conditions:

• one or more hospitalizations with a diagnosis for depressive disorder, affective psychoses, neurotic depression or adjustment reaction, ICD-9-CM codes 296.2-296.8, 300.4, 309, 311; ICD-10-CA codes F31, F32, F33, F34.1, F38.0, F38.1, F41.2, F43.1, F43.2, F43.8, F53.0, F93.0; OR
  
  
• one or more physician visits with a diagnosis for depressive disorder, affective psychoses or adjustment reaction, ICD-9-CM codes 296, 309 or 311; OR
  
  
• one or more hospitalizations with a diagnosis for anxiety disorders, ICD-9-CM code 300; ICD-10-CA codes F32.0, F34.1, F40, F41, F42, F44, F45.0, F45.1, F45.2, F48, F68.0, F99, AND one or more prescriptions for an antidepressant or mood stabilizer, ATC codes N03AB02, N03AB52, N03AF01, N05AN01, N06A; OR
  
  
• one or more physician visits with a diagnosis for anxiety disorders, ICD-9-CM code 300, AND one or more prescriptions for an antidepressant or mood stabilizer, ATC codes N03AB02, N03AB52, N03AF01, N05AN01, N06A

3. Martens et al. (2010)

In the Health Inequities deliverable (Martens et al., 2010), depression was defined as residents aged 10 or older over a five-year period with any one of the following conditions:

• one or more hospitalizations with a diagnosis for depressive disorder, affective psychoses, neurotic depression or adjustment reaction, ICD-9-CM codes 296.2-296.8, 300.4, 309, 311; ICD-10-CA codes F31, F32, F33, F34.1, F38.0, F38.1, F41.2, F43.1, F43.2, F43.8, F53.0, F93.0; OR
  
  
• one or more physician visits with a diagnosis for depressive disorder, affective psychoses or adjustment reaction, ICD-9-CM codes 296, 309, 311; OR
  
  
• one or more hospitalizations with a diagnosis for anxiety disorders, ICD-9-CM code 300; ICD-10-CA codes F32.0, F34.1, F40, F41, F42, F44, F45.0, F451, F452, F48, F68.0, F99; OR
  
  
• one or more physician visits with a diagnosis for anxiety disorders, ICD-9-CM code 300.

NOTE: In the Health Inequities deliverable (Martens et. al, 2010), the definitions for identifying cumulative mental illness (of which depression is one of five conditions) did not include prescriptions from the Drug Program Information Network (DPIN) data. The reason for this is to maintain consistency in the definition across time, because the earliest analysis time frame (1984) pre-dated the availability of the DPIN data (1995). For an illustration of the difference of the crude rate for cumulative mental illness with and without DPIN (drug) data over time, please see the Cumulative Mental Illness glossary term.

4. Katz et al. (2014)

In the Physician Integrated Network: A Second Look deliverable by Katz et al. (2014), they investigated depression care related to the implementation of Physician Integrated Network (PIN) clinics. Depression care was calculated as the percent of patients newly diagnosed with depression, as well as those who filled a prescription for an antidepressant medication and made three subsequent ambulatory visits within four months of the prescription being filled (any diagnosis, any physician).

Depression was defined as:


• one physician visit with a diagnosis for depression: ICD-9-CM codes 296 or 311 AND
  
• who filled a prescription for an antidepressant medication (ATC code N06A) within two weeks of the diagnosis
  
  NOTE: To be included as a newly depressed patient, there could not be a previous physician visit with a diagnosis of depression or a prescription for antidepressants in the two years prior to the current diagnosis.

For more information, please see the section titled Depression Care in this deliverable.

Issues and Notes

Eilish Cleary ran comparisons using NPHS and various indicators of depression. While there were similar rates of depression and depression related measures for one year rates in the NPHS and administrative data the actual individuals were not the same. Rates of depression seen the CCHS and other surveys also have similar rates. Before using the administrative definition for identifying actual individuals, please review the appropriate section in the Mental Health deliverable. Although a population prevalence estimate is relatively similar no matter what the source - survey data or administrative claims data - researchers must be cautious when attributing the diagnosis to an individual person. In other words, studies which propose to study the individual effects of depression may be problematic.

Bipolar illness is defined in ICD-9-CM at the fourth-digit level under the 296 code known as Episodic Mood Disorder, and therefore can only be identified in the hospital discharge abstract data. However, because diagnoses in the physician claims data are only recorded at the third-digit level, it is not possible to specifically identify bipolar disorder in the physician claims data.

The mental health deliverable looked only at the existence of depression and drug codes within a 5 year period. Further work in the child inequities project is looking at maternal depression. Currently there is no temporal (or time dependence) with regard to the use of drugs and/or diagnosis.

The mental health deliverable included MHMIS data as part of the definition. While this added some individuals the number was not great. Special permission is required to use the MHMIS data and in most studies it will not be necessary.

There have been a variety of definitions using ICD-9-CM codes in the literature. The June 2004 volume of Medical Care had a number of articles on depression that provide an idea of the variation in definitions that have been used. Four of these articles use ICD-9-CM codes at the four character level. Himelhoch et al. used 296.2-296.3, 300.4, 309.0, 311 having either an inpatient or 2 outpatient claims within a single year. Charbonneau et al. used 296.2 or 296.3, 311 during a two year period and at least two antidepressant from a VHA pharmacy during the study period. Busch et al. used only a single code of 296.2 or 296.3 and who received antidepressant medications from a VA hospital over a one year period. Hamed et al. over a two year period used at least one of the following 296.20-296.26, 296.30-296.36, 300.4, 311.0, and at least 1 prescription for TCAs, SSRIs, SNRIs, and not a single code for bipolar (296.0, 296.1, 296.4-296.9) or schizophrenia (295). It should be noted that all of these studies were looking for a specific cohort of individuals to study and were not looking at depression in the general population.