How Do Delays in Mental Health Treatment Impact Long-Term Mortality?

The research examines how delays in mental health treatment impact long-term mortality, particularly among veterans experiencing mental health emergencies. Using data from over 621,000 VA emergency department (ED) visits between 2001 and 2015, the study constructs an exogenous measure of clinic congestion to isolate causal effects. Findings show that a one standard deviation (11.7-day) increase in wait time raises two-year mortality by 1.5%. The effect is driven by disengagement from mental health care, particularly among patients with substance use disorders and psychosis. The study highlights the critical role of timely mental health care in reducing long-term mortality risks.

Full Citation and Link to Article

Costantini, Sydney. "How Do Mental Health Treatment Delays Impact Long Term Mortality?

AMERICAN ECONOMIC REVIEW (FORTHCOMING). https://www.aeaweb.org/articles?id=10.1257/aer.20240226&&from=f

Extended Summary

Central Research Question

This study investigates how delays in mental health treatment affect long-term mortality, particularly among veterans who visit emergency departments (ED) with mental health-related conditions. With a growing mental health crisis and long wait times due to clinic congestion, many patients face challenges in accessing timely care. The research aims to isolate the causal impact of wait times on mortality, asking: Does delayed mental health treatment increase long-term mortality?The study leverages data from the Veterans Affairs (VA) healthcare system, where treatment is free or low-cost, making it an ideal setting to examine the effects of access barriers unrelated to financial constraints.

Previous Literature

Previous research highlights the link between mental health care access and patient outcomes. Studies have shown that engagement with mental health services can improve employment outcomes, education, and overall quality of life (Biasi et al., 2020; Bhat et al., 2022). Other work has demonstrated the impact of mental health care on reducing suicide rates (Lang et al., 2013; Feyman et al., 2022). However, most studies do not examine long-term mortality as a primary outcome.

A closely related study by Feyman et al. (2022) finds that reductions in mental health staffing at the VA lead to increased suicide rates among veterans. This suggests that barriers to mental health care—including clinic congestion and long wait times—may have severe consequences. Additionally, research on wait times for physical health conditions shows that treatment delays can negatively impact outcomes, particularly for time-sensitive conditions such as cancer and cardiovascular disease (Reichert & Jacob, 2018; Singh & Venkataramani, 2022). This study builds on these findings by focusing on how mental health treatment delays affect mortality, using a unique empirical approach that isolates the effects of congestion-driven wait times.

Data

The study utilizes administrative data from the VA’s Corporate Data Warehouse, which contains comprehensive patient records, including all emergency department visits, mental health appointments, diagnoses, and mortality data. The dataset includes 621,289 veterans who visited a VA ED for a mental health-related condition between 2001 and 2015.

Patient demographics include age, race, ethnicity, and gender, along with prior healthcare utilization history. The primary outcome variable is two-year mortality, which is measured using data from the Veterans Benefits Administration (VBA), Centers for Medicare and Medicaid Services (CMS), and Social Security Administration (SSA), ensuring a highly accurate measure of patient death. The study also examines intermediate outcomes, such as attendance at follow-up mental health appointments, long-term engagement with therapy, and prescription drug use.

To measure mental health treatment delays, the author constructs “mean wait time”—the average wait time for all other patients who visited the same hospital in the same two-week period and were scheduled for the same type of mental health appointment (e.g., psychiatry, psychology, substance use disorder clinic). This approach ensures that the study captures clinic congestion rather than individual patient scheduling preferences or risk levels.

Methods

A key challenge in studying the effect of mental health treatment delays on mortality is isolating causality—patients with longer wait times may differ systematically from those with shorter waits. To address this, the study employs a quasi-experimental design using leave-out mean wait time as an exogenous measure of congestion.

The main regression model estimates the effect of wait time on mortality while controlling for:

Hospital fixed effects (to account for differences in hospital quality)
Year-month fixed effects (to adjust for seasonal or policy-driven changes)
Appointment type fixed effects (to ensure comparisons are made within similar types of treatment)
Demographics and health history (to control for observable patient risk factors)

To ensure that wait times are randomly assigned within hospital-time-appointment type groups, the study tests whether predicted mortality (based on patient characteristics) is correlated with assigned mean wait time. The results confirm that high-risk and low-risk patients are equally likely to be exposed to long wait times, supporting the assumption of quasi-random assignment.

Additional robustness checks include:

Removing patients with non-VA insurance, to rule out the possibility of patients seeking mental health care outside the VA system
Using an alternative predicted wait time measure to address concerns that appointment type assignments may be endogenous
Examining placebo outcomes, such as mortality for patients who visit the ED for circulatory conditions, to ensure that mental health wait times do not simply reflect overall hospital congestion

Findings/Size Effects

The study finds a significant and robust relationship between mental health treatment delays and long-term mortality. Specifically:

A one standard deviation (11.7-day) increase in mean wait time raises two-year mortality by 1.5%.
The effect persists for at least five years after the initial ED visit, suggesting lasting consequences.
The impact is strongest for patients with substance use disorders (SUD) and psychosis, who are already at higher risk of premature death.
Longer wait times make patients less likely to attend their follow-up mental health appointment. A one standard deviation increase in wait time reduces follow-up attendance by 2.0 percentage points (5.6% relative to the mean).

The study also explores potential mechanisms driving this effect:

Missed Appointments & Disengagement from Care
- Patients with longer wait times are more likely to disengage entirely from mental health care—both outpatient and inpatient treatment.
- Missing the first follow-up appointment increases the risk of long-term disengagement, reinforcing the importance of timely scheduling.
Compensatory Behavior by Physicians
- ED doctors appear to anticipate longer wait times and compensate by prescribing more psychiatric medications at the initial ED visit.
- A one standard deviation increase in wait time raises the probability of receiving a prescription by 4.0%.
- Despite this, medication alone does not offset the long-term mortality effect, underscoring the importance of continued engagement with therapy.
Cause of Death Analysis
- The majority of excess deaths among patients with long wait times are due to physical conditions, rather than suicide or overdose.
- Specifically, patients with longer wait times have higher mortality from heart disease, respiratory diseases, liver disease, and infections—conditions that are known to worsen with untreated mental illness.

These findings align with prior research showing that patients with severe mental illness die 10–20 years earlier than the general population, often due to preventable physical health conditions (Liu et al., 2017).

Conclusion

This study provides compelling evidence that mental health treatment delays significantly increase long-term mortality, primarily by reducing follow-up appointment attendance and long-term engagement with therapy. The findings suggest that seemingly minor scheduling delays can have major consequences for high-risk populations, particularly veterans with substance use disorders and psychosis.

The results have important policy implications:

Reducing wait times for mental health appointments could save lives, particularly for high-risk patients.
Efforts to increase mental health staffing and appointment availability should be prioritized, especially in overburdened VA hospitals.
ED physicians’ compensatory behaviors (such as increased psychiatric prescribing) highlight the need for integrated mental health services that ensure continuity of care beyond the ED.

By demonstrating that even modest delays in care can drive long-term disengagement and higher mortality, this study underscores the urgent need for systemic improvements in mental health service delivery. Addressing treatment delays could be a cost-effective and life-saving intervention for vulnerable populations.