• How do acute physical health shocks impact subsequent mental health service utilisation among existing users of specialist NHS mental health care?


• Population: Users of specialist mental health care


• Exposure: Acute physical health shocks (AMI, stroke, cancer diagnosis)


• Outcome: Mental health service utilisation

• NHS England provides primary, secondary, and specialist mental health care.


• Specialist mental health services primarily delivered by NHS Mental Health Trusts.


• High rates of physical and mental health comorbidities among patients.

• Existing mental health conditions can exacerbate the impact of acute physical health shocks.


• Increased risk of disruptions in continuity of mental health care following a physical shock.


• Understanding utilisation patterns is crucial to inform targeted care integration.

Analysis uses linked NHS administrative data from two main sources:

• Hospital Episode Statistics (HES): Acute inpatient admissions data
• Mental Health Services Data Set (MHSDS): Specialist mental health care data

• Primary care (GP and IAPT referrals)


• Community-based mental health services


• Secondary and tertiary care (NHS Mental Health Trusts)


• Integrated care within physical health settings

• Propensity score matching based on demographics and health status


• Matching variables: Age, Sex, IMD quartile, Elixhauser comorbidity index, Mental health clusters


• Comparison group: Mental health users without acute physical health shocks during study period

• Acute physical health shocks identified via first recorded events:


• Heart attack (AMI)


• Stroke (Cerebrovascular accident)


• Cancer diagnosis




• Event timing precisely established from hospital admission records

• Primary outcome: Days spent in specialist mental health care post-shock


• Tracked quarterly to observe short- and medium-term impacts


• Includes inpatient, outpatient, and community care settings

• Demographic factors: Age groups (18-35, 36-55, 56-75, 76+), Sex, Ethnicity (White, Black, Asian, Mixed, Other), IMD quartiles (most deprived to least deprived)


• Clinical characteristics: Elixhauser comorbidity index, Mental health clusters (psychosis, non-psychosis, dementia, and others)


• Ensures comparability of matched cohorts

\( y_{it} = \alpha_0 + \alpha X_i + \beta d_i + \gamma \lambda + \color{orange}{\delta d_i \lambda} \) + \( u_{it} \)

Each \( \lambda \) represents a three-month period

\( y_{it} = \alpha_0 + \alpha X_i + \beta d_i + \sum\limits_{t=2}^{T} \gamma_t \lambda_t + \sum\limits_{t=T_s}^{T} \delta_t(d_i \lambda_t) + u_{it} \)

• Covariate balance assessed pre- and post-matching for inpatient cohort
• Standardized mean differences below 0.1 indicate good balance
• Ensures comparability between treatment and control groups

• Total number of inpatient bed-days: 7.4m for 2016/17, 7.5m for 2019/20
• Average number of inpatient bed-days: 7 to 8 days per quarter
• Higher utilisation in July to December time

• Covariate balance assessed pre- and post-matching for outpatient cohort
• Standardized mean differences below 0.1 indicate good balance
• Ensures comparability between treatment and control groups

• Total number of days spent in community care: 151m 2016/17, 135m 2019/20
• Average number of days range from 20 to 25 days per quarter

• Acute physical health shocks substantially alter mental health service use.


• Impact varies significantly by shock intensity, highlighting treatment heterogeneity.


• Short-term disruptions evident; longer-term trends merit exploration.

• Highlight the need for integrated physical-mental healthcare strategies post-shock.


• Potential role for liaison psychiatry services in acute care settings.


• Policy recommendations to enhance continuity of care across NHS services.

• Robust matched cohort design and multi-timepoint DiD framework.
• Choice of stacked DiD over staggered DiD for clearer temporal alignment of shocks.
• Analysis focuses on first-ever acute physical health shocks; repeat events not captured.
• Chronic nature of some conditions (e.g., cancer) may lead to underestimation of impact.
• Findings valid within the context of secondary mental health care users in NHS England.

• Highlight potential increased NHS resource utilisation post-shock.


• Economic justification for targeted preventative mental health interventions.


• Clinical emphasis on integrated care improving long-term patient outcomes.

• Exploring longer-term mental health trajectories post-acute health shocks.


• Assessing mental health outcomes in primary care and informal caregiving contexts.


• Evaluating integrated interventions designed to mitigate mental health impacts.