Table 1 Study characteristics and modelling strategies
From: System dynamics models of depression at the population level: a scoping review
Patten 2002 [26] | Tam 2020a [27] | Tam 2020b [28] | Tandon 2021 [29] | |
|---|---|---|---|---|
Purpose | Evaluate the impact of antidepressant medication use and long-term preventive treatment on the prevalence of MDE | Evaluate the impact of recall error on the prevalence of lifetime MDE | Evaluate trends in smoking, smoking-attributable mortality, and life-years lost by MDE status. Evaluate the impact of smoking cessation among adults with depression | Estimate the impact of counselling rates and transmissibility of depression through social connections on depression estimates |
Population | Canada general population | USA general population | USA population aged 18 + | Zimbabwe general population |
Definition of depression | Past-year MDE using the CIDI-SF. Current MDE as past-year MDE and current psychological distress | Lifetime and past-year MDE using DSM-IV criteria | See Tam 2020a | Current depression defined as an extreme case of negative emotion |
Depression states | • Never depressed, first, second, or third/subsequent MDE • First, second, and third episodes of remission • Short duration/good prognosis, intermediate duration/prognosis with antidepressant use, intermediate duration/prognosis without antidepressant use, long duration/poor prognosis | • Never reported MDE • Past-year MDE • Former MDE • Former MDE with recall error | MDE states (see Tam 2020a) among people who have never smoked, people who smoked previously, people who currently smoke | • No depression • Primary depression, mild or initial stage • Secondary depression, severe or acute stage • Recovered depression |
Transitions between depression states (i.e. flows) | • Incidence of first MDE • Remission of first, second, and third/subsequent MDE • Recurrence of second and third/subsequent MDE from first, second, or third/subsequent MDE | • Incidence of past year MDE • Recovery from past year MDE to former MDE • Incidence of underreporting from former MDE to recall error • Recurrence of past year MDE from former MDE or recall error | See Tam 2020a among adults who are current, former, or never smokers | • Incidence of primary depression • Recovery of primary depression • Incidence from primary to secondary depression • Recovery of secondary depression |
Feedback loops | • Feedback loop between incidence of third/subsequent MDE and recovery of third/subsequent MDE | • Feedback loop between former MDE and past year MDE • Feedback loop between former MDE, recall error, and past year MDE | See Tam 2020a among different categories of smoking status | • Number of people with secondary depression creates feedback loop by influencing incidence rate of primary depression |
Data used for depression and parameter estimates | • National epidemiologic studies (National Population Health Survey, 1996 cycle; Stirling County study) • Literature reviews | • National epidemiologic studies (Epidemiologic Catchment Area Study) • Calibrated estimates | See Tam 2020a | • Unknown sources for fixed parameters • Calibrated estimates |
Calibration data | National Population Health Survey 1996 cycle | National Survey on Drug Use and Health 2017, Baltimore ECA study 1981–2005 | See Tam 2020a | Annual report of cases of depressed population in Zimbabwe from 1991 to 2017 |
Interventions or policy changes | Increasing antidepressant use among people with intermediate prognosis of MDE Changes in recurrence rates for people with three or more MDEs though long-term treatment | Maximum potential reduction in premature mortality scenario, where all smokers with MDE quit immediately in 2018 and no new smoking initiation occurs | Increase in rate of counselling among people with primary and secondary depression | |
Is model replicable | Yes | Yes | Yes | No |