Gronholm, Petra C;
Knapp, Martin;
Brimblecombe, Nicola;
Maughan, Barbara;
Richards, Marcus;
Bonin, Eva-Maria;
Hoffmann, Mauricio S;
... Evans-Lacko, Sara; + view all
(2022)
Health Service Costs in Adulthood Associated with Adolescent Mental Health Problems in Three British Cohorts.
Journal of Mental Health Policy and Economics
, 25
(2)
pp. 39-52.
![[thumbnail of Richards_LT econ impacts of adolescent MH probs_ JMHPE_manuscript_R1_220422_CLEAN (1).pdf]](https://discovery.ucl.ac.uk/style/images/fileicons/text.png) |
Text
Richards_LT econ impacts of adolescent MH probs_ JMHPE_manuscript_R1_220422_CLEAN (1).pdf Access restricted to UCL open access staff Download (512kB) |
Abstract
BACKGROUND: Mental health problems are common among young people and they can have long lasting effects into adulthood. These effects can have economic consequences, both in the short term and into adulthood and across the life course. AIMS OF THE STUDY: To present a methodology to estimate harmonised costs across three large-scale British birth cohorts (1946 Medical Research Council National Survey of Health and Development (NSHD); 1958 National Child Development Study (NCDS); and 1970 British Cohort Study (BCS70) for key health service use measures: restrictive and/or urgent care (including inpatient, day surgery and accident and emergency services), outpatient services (including mental health specialist), and general practitioners (GP). We use this to describe relationships between adolescent mental health (assessed at ages 13/15 for NSHD, and age 16 for NCDS and BCS70) and the economic impacts associated with health service use in early- and mid-adulthood. This methodology can be used to explore long-term economic outcomes associated with emotional and behavioural problems experienced by adolescentsin other longitudinal cohort studies. METHODS: For each cohort, we analysed data on participant reports of health service use. We categorised services into broad domains and estimated frequency of service use at comparable timepoints around ages 20, 30 and 40 (NSHD ages 26, 31/36, 43; NCDS ages 23, 33, 42; BCS70 ages 26, 30, 42), by adolescent mental health status (mild and severe emotional problems; mild and severe conduct problems; absence of problems) and sex. Data collection methods varied between cohorts and across time-points, and data on health service use frequency were not always comprehensively collected at each wave. Where frequencies were unavailable, for each cohort we estimated frequency of use from contemporaneous resources and applied conservative assumptions so economic impacts would not be overestimated. We then estimated service use costs for each participant in the different cohorts based on their estimated frequency of health service contact multiplied by relevant unit costs. RESULTS: Our findings illustrate the capacity to conduct coordinated analyses across three different British birth cohorts and identify patterns of service utilisation for respondents around the ages of 20, 30 and 40 years of age. Our analyses identified similar patterns of utilisation for respondents at around age 20, 30 and 40. For instance, those with conduct problems, on average, had about 1.5 times higher costs associated with GP service use at age 43 compared to those with emotional or no mental health problems. DISCUSSION: This study demonstrates the potential for comparing service use and associated costs by age, type of health service contact and adolescent mental health problem across different birth cohorts. Our methodology provides a resource that enables examinations of associations between adolescent mental health problems and health service use in adult life, as well as the associated costs, across three British birth cohort datasets. IMPLICATIONS FOR HEALTH POLICIES: Insights from this research could provide methodological and empirical bases to support further work on adolescent mental health, for example to examine the case for shifting resources to early intervention and prevention. IMPLICATIONS FOR FURTHER RESEARCH: Future research could investigate in more detail changes in economic impacts over the life course among other population subgroups. Moreover, when combined with information on costs and outcomes of interventions, estimates could be used to help identify interventions with the potential to provide good value for money.
Archive Staff Only
 |
View Item |
