Promoting suicide prevention in hospitals: Evidence and incentives

Executive summary

Suicide is one of the leading causes of death in the United States, with over 49,000 deaths by suicide in 2023 alone, a trend that shows no sign of abating. Many of these deaths occurred within reach of the health care system: approximately 13 to 30% of individuals who die by suicide had an inpatient stay or emergency department visit in the month before their death.¹ Research has identified four core practices through which hospitals can reduce this risk—formal safety planning, lethal means counseling, warm handoffs to outpatient care, and follow-up contact after discharge—each with demonstrated potential to reduce suicidal behavior and improve care transitions. Yet only 8% of accredited hospitals have implemented all four concurrently. This gap reflects neither a failure of clinical knowledge nor a lack of evidence, but a misalignment of incentives: Hospitals bear the costs of implementation, while the financial benefits accrue largely to insurers. 

Payers, therefore, hold the leverage to close this gap. Health plans and payments overseen by the Centers for Medicare and Medicaid Services (CMS) alone account for nearly half of hospital revenues, a share large enough to move hospital behavior nationwide and one that reaches the resource-constrained hospitals least able to fund these practices on their own.  

We propose fidelity-based quality process metrics for each of the four recommended practices—assessing the timing and completeness of safety planning and lethal means counseling, administrative readiness and completion of warm handoffs, and the timing, modality, and consistency of follow-up contacts—rather than metrics tied to downstream clinical outcomes.  

Payment should then be linked to these measures in a way that rewards both progress and performance: Hospitals would earn payments for improving, but the largest rewards would be reserved for reaching meaningful standards of care, with higher payments for the practices that are hardest to implement because they extend beyond the hospital’s walls. Because hospitals begin at widely different levels of capacity and adoption, these incentives should be introduced gradually. Payers should start with bonus-only payments that build participation and infrastructure, later transitioning to a two-sided risk model in which hospitals equipped to take on that risk face financial consequences for underperformance alongside bonuses for strong performance. Both stages draw on approaches already in use in Medicare through the Center for Medicare and Medicaid Innovation (CMMI) and in state Medicaid programs, such as Colorado’s.

Introduction

Suicide is one of the leading causes of death in the United States, with over 49,000 deaths by suicide in 2023 alone, a trend that shows no sign of abating. Reducing these deaths is a recognized public health priority, and interventions have been studied across the care continuum to identify the most effective points of impact. Hospitals are particularly well-positioned to help, given that approximately 13 to 30% of individuals who die by suicide had an inpatient stay or emergency department visit in the month before their death, and 36 to 86% in the year before their death.²

Hospitals represent one critical, but underutilized, intervention point in the continuum of suicide prevention. Emergency department (ED) visits for suicide attempts and intentional self-harm nearly quadrupled between the early and late 2010s, rising from 1.43 million to 5.37 million over the course of the decade, and individuals who present following a suicide attempt face a 30-times greater risk of subsequent suicide death. Even among those with a known suicide attempt, only 55% received any inpatient or outpatient mental health care in the same year. Hospitals are thus a critical setting for high-risk individuals, presenting after a suicide attempt, self-inflicted injury, or suicidal crisis, and potentially, those whose risk has not yet been identified. 

Despite this, hospitals have been slow to adopt evidence-based practices that could meaningfully reduce harm during and after these encounters. Universal suicide risk screening is widely recommended, and roughly two-thirds of Joint Commission–accredited hospitals report hospital-wide screening protocols. However, adoption of the full complement of recommended practices remains to be fully embraced. A growing body of research and clinical guidance, complemented by the Zero Suicide framework and the Joint Commission’s National Performance Goals (NPGs), has identified four core practices for hospital-based suicide prevention:  

1. formal safety planning, 
2. lethal means counseling,  
3. warm handoffs to outpatient care, and  
4. follow-up contact after discharge. 

Each practice has demonstrated the potential to reduce suicidal behavior and improve care transitions. Yet only 8% of accredited hospitals responding to the Joint Commission’s 2022 survey had implemented all four practices concurrently. 

The economic realities of how hospitals are paid and how quality improvement is financed limit hospitals’ ability to adopt best practices in suicide prevention. Quality improvement is costly, and the primary financial benefits of better suicide prevention and care accrue largely to insurers rather than to the hospitals bearing the costs of implementation, for example, through reduced readmissions or fewer return ED visits. This misalignment of incentives appears to have constrained the voluntary adoption of these suicide prevention practices in hospitals and suggests that goodwill and clinical guidance alone are insufficient levers to drive systemic change. Yet some hospitals continue to provide this care despite the absence of financial benefit, reflecting health system leaders’ enduring commitment to patient safety. That commitment should be applauded, but it clearly does not motivate the majority of hospitals to adopt the recommended practices. Well-designed financial incentives can align a hospital’s economic interest with evidence-based patient care.  

The leverage to create such incentives is challenging, but the potential exists across the payer landscape. Commercial insurers, state Medicaid programs, and the federal government all hold financial relationships with hospitals that could be tied to suicide prevention performance—and health plans and payments overseen by the Centers for Medicare and Medicaid Services (CMS), spanning Medicare, Medicaid, and the Affordable Care Act (ACA) Marketplaces, alone account for nearly half of hospital revenues, a share large enough to affect hospital behavior nationwide. 

This analysis explores how payers and hospitals can be engaged as active partners in promoting hospital adoption of suicide prevention best practices through fidelity-based quality metrics and payment arrangements that create real financial incentives for implementation. Rather than focusing on any single payer type, we examine what hospitals should be doing, why current incentive structures have failed to get them there, and what payers could do to close that gap. 

The paper is organized into the following sections: We first provide background on the best practices in hospital-based suicide prevention and hospitals’ fidelity to them. We outline the economic considerations shaping hospitals’ engagement with these practices, followed by a discussion of quality metrics and payment strategies to incentivize hospital participation. We close with a brief comment on the role of health plans and the federal government as payers. 

Background

Efforts to standardize suicide prevention in health care settings have increasingly coalesced around a common framework called Zero Suicide, and from that framework, a specific set of evidence-based practices has emerged as the core of what hospitals should be doing in their EDs and inpatient units to reduce suicide risk before, during, and after discharge.

Best practices in hospital-based suicide prevention

The Zero Suicide framework

The Zero Suicide framework emerged from a partnership between the National Action Alliance for Suicide Prevention, the Suicide Prevention Resource Center, and other national suicide prevention experts, inspired by the Perfect Depression Care initiative at Henry Ford Health System in Michigan in the early 2000s. The Perfect Depression Care initiative produced notable results: Henry Ford achieved an 80% reduction in suicides, including an 18-month period between 2009 and 2010 with zero suicides—a statistically significant decrease that occurred even as Michigan’s overall suicide rate climbed. 

Zero Suicide became the first and most widespread model for quality improvement in suicide prevention at the health care organization level, targeting system-wide cultural and structural change rather than individual clinical behavior. Its foundational premise is that suicide deaths among patients receiving care are preventable, and that achieving that goal requires a “whole-organization” commitment: standardized, evidence-based behavioral health protocols, a trained workforce, and continuity of care across the entire process of treatment and discharge. The framework has since been adopted as a guiding model by the Substance Abuse and Mental Health Services Administration (SAMHSA), has informed the Joint Commission’s NPGs on suicide prevention, and is now being implemented, to varying degrees, across a growing number of health care organizations in behavioral health programs, general and psychiatric hospitals, primary care settings, and health plans. 

Zero Suicide operationalizes this commitment to suicide prevention in health care systems through seven elements—Lead, Train, Identify, Engage, Treat, Transition, and Improve—that together address both administrative infrastructure and direct clinical care. This analysis focuses on one element, Transition, or discharge planning, which encompasses the four evidence-based practices noted earlier. These practices are relevant across the full continuum of suicide prevention care but are particularly critical at points of highest acute risk, including discharge from inpatient psychiatric units and EDs following a suicide attempt or crisis. Gaps in hospital implementation are well-documented and are the focus of this analysis. 

Four evidence-based protocols in suicide prevention

Four practices sit at the center of this analysis: safety planning, lethal means counseling, warm handoffs to outpatient care, and follow-up contact. Recommended by The Pew Charitable Trusts and the Joint Commission and grounded in the Zero Suicide framework, each has a substantial evidence base supporting its effectiveness in reducing suicidal thoughts and behavior, yet most hospitals consistently underdeliver on each. The sections below review the evidence for each practice and document the gap between what research recommends and what hospitals currently provide—a gap that structured payment incentives could help close.

Safety planning

Safety planning is a brief, collaborative intervention in which a clinician and patient work together to create a prioritized, written set of coping strategies and sources of support to be used during and after a suicidal crisis. It has become one of the most widely recommended brief interventions for individuals identified as at risk across inpatient, emergency, and outpatient settings. Safety plans incorporate elements of several evidence-based suicide risk reduction strategies, including access to lethal means reduction, brief problem-solving and coping skills, social and emergency crisis support, and motivational enhancement for continuing behavioral health treatment. A 2022 systematic review of 26 studies across general adult and veteran populations found that these safety planning interventions were associated with improvements in suicidal ideation and behavior, decreases in depression and hopelessness, reductions in hospitalizations, and improvements in treatment attendance, with qualitative studies further confirming their acceptability and feasibility across a range of clinical settings. 

Research shows that higher-quality safety plans, administered by trained staff and tailored to the individual at risk, are associated with reduced likelihood of hospitalization in the year following safety planning. Yet studies have found that safety plans in practice are frequently incomplete or insufficiently personalized. One inpatient study found that mandating safety plans without accompanying training and implementation support was not sufficient to ensure quality. 

Safety planning is also most effective when paired with follow-up during and after discharge that updates the plan, supports engagement with outpatient care, and identifies social contacts that a patient can reach out to when experiencing suicidal ideation after the visit—a combination that has been shown to decrease the likelihood of subsequent suicide attempts and inpatient psychiatric hospitalizations.

Lethal means counseling

Lethal means counseling consists of a conversation between a clinician and a patient, and ideally their support network, aimed at reducing access to firearms, medications, and other highly lethal means during periods of elevated suicide risk. It can be administered in under five minutes by any trained health care worker. The intervention may be incorporated into safety planning or delivered on its own. However, while lethal means reduction is embedded as a core component of safety planning, it warrants particular attention as a clinical practice in its own right, given the strength of the supporting evidence and the high frequency with which it is omitted or insufficiently addressed in safety plans. 

The case for lethal means counseling rests on two intersecting bodies of evidence: that restricting access to lethal means reduces suicide rates at the population level, and that the window between suicidal crisis and attempt is often extremely short. International studies have demonstrated that overall suicide rates drop when access to commonly used, highly lethal methods is reduced. Among survivors of nearly lethal suicide attempts, 24% reported taking less than five minutes between the decision to attempt and the actual attempt. Additionally, studies consistently find that the interval between the emergence of suicidal thoughts and an attempt can be ten minutes or less, leaving little time for intervention once a crisis has begun. Reducing access to lethal means during a moment of acute crisis can therefore be lifesaving, and evidence suggests that if a favored method becomes less available, individuals do not reliably substitute another method. 

Warm handoffs to outpatient care

The transition from inpatient or emergency care to outpatient behavioral health treatment is among the highest-risk periods for individuals who have experienced a suicidal crisis. In the first week after discharge from inpatient psychiatric care, suicide death rates are approximately 300 times higher than that of the general population and may remain elevated for up to three months post-discharge. Rather than providing a referral contact at discharge and leaving individuals to navigate the barriers of outpatient care on their own, a warm handoff directly connects the patient with their next provider before discharge and establishes a relationship with the receiving provider while the patient is still in a supported care environment. Research supports the importance of this transition: brief interventions that include care coordination have been shown to reduce subsequent suicide attempts for individuals who presented with suicidal thoughts or behaviors in an ED or other medical setting. Separately, one study examining Medicaid-enrolled youth across 33 states found that timely follow-up treatment that can be promoted by warm handoff interventions—specifically, seeing a mental health provider within seven days of discharge from inpatient psychiatric care—was associated with meaningfully reduced suicide risk. The evidence for warm handoffs is compelling and backed by a large body of literature showing that patients who receive one, along with other transitional interventions, are significantly more likely to attend their first outpatient appointment or follow-up treatment. The need for this practice to improve continued patient engagement with care is reinforced by data on post-discharge follow-up: fewer than half of patients discharged from EDs for a mental health crisis successfully transition to outpatient care, with high rates of ED readmission.  

A warm handoff directly targets this gap by reducing the logistical and psychological barriers between discharge and first outpatient contact. The National Action Alliance for Suicide Prevention’s best practices guidance recommends that inpatient providers secure an outpatient appointment for their patient within 24 to 72 hours of discharge, initiate direct personal contact between the patient and the receiving outpatient provider before discharge, and proactively work with the patient and family to identify and resolve logistical barriers such as transportation, childcare, and insurance before the patient leaves the inpatient setting. 

Follow-up contact after discharge

Follow-up contact involves reaching out to individuals by phone or other means within the days and weeks following hospital discharge to assess ongoing risk, offer support, and help address barriers to accessing outpatient care. Follow-up contact, ideally continuing until the patient attends their first outpatient appointment, is supported by evidence as a component of reducing subsequent suicide attempts and improving treatment engagement. A collection of research indicates that active follow-up contact interventions are effective in reducing the risk of a repeat suicide attempt within six months. 

The National Action Alliance for Suicide Prevention’s best practices guidance recommends a follow-up call within 24 to 48 hours of inpatient discharge, weekly telephone contact until the patient attends their first outpatient appointment, and ongoing caring contacts—brief, encouraging messages to the patient that do not require a response. Outpatient providers should reach out to confirm the first appointment if it is more than 24 hours after discharge, and written protocols should govern follow-up on missed appointments, escalating from telephone contact to mobile crisis outreach and, where risk indicators warrant, emergency welfare checks. Research supports extending these contacts over longer periods: one study found that the rate of suicide attempts was 33% lower among individuals who received up to three follow-up contacts compared with those who received none. 

Follow-up contact should also help ensure that patients maintain contact with health care providers and support promptly attending a scheduled outpatient appointment. A large population-based cohort study using data of over 76,000 patients discharged from psychiatric inpatient care in South Korea found that earlier outpatient appointments were significantly associated with lower suicide risk, with patients seen within seven days showing meaningfully reduced risk across diagnoses, including depression, schizophrenia, and bipolar disorder, relative to patients not receiving care within 30 days. A combined safety planning and structured follow-up contact intervention tested across nine Veterans Affairs EDs was associated with 45% fewer suicidal behaviors over a six-month follow-up period and more than double the odds of attending at least one outpatient mental health visit compared with usual care. A review of eleven empirical studies found that repeated follow-up contacts are associated with reductions in suicidal behavior. 

Taken together, these four practices form an important part of an evidence-based continuum of care that addresses suicide risk not as a single clinical event but as a sustained process requiring attention from the moment of identification through the critical weeks following discharge. Yet despite this clear evidence base, hospitals have struggled to implement these practices consistently and at scale. This implementation gap is not due to a failure in clinical knowledge or evidence but largely reflects the structural and financial realities hospitals navigate. Hospitals bear real costs in training staff, embedding protocols, and sustaining follow-up infrastructure, while the financial benefits of better suicide prevention care accrue largely to insurers through reduced utilization. The data on current hospital practices make this misalignment concrete. 

Hospital fidelity

Existing evidence suggests that about 13 to 30% of individuals who die by suicide had an inpatient stay or ED visit in the month before their death; moreover, in the year before death, these figures rise to about 36 to 86% of individuals.³ It is for this reason that increasing health system fidelity to the use of evidence-based practices aimed at suicide prevention is so important. 

According to the Joint Commission, suicide prevention best practices in hospital settings should include suicide screening, formal safety planning, warm handoff to outpatient care, follow-up with patients after discharge, and lethal means counseling. In several recent studies, the Joint Commission collected data on its accredited hospitals to better understand how often providers are engaging in these practices.⁴ Overall, suicide screening is the most prevalent practice, with about 80% of hospitals reporting universal screening and 65% of hospitals having implemented protocols for positive screens and risk assessments. However, this still leaves about one in five hospitals that do not screen all patients, most often due to insufficient resources to adequately address patients who ultimately screen positive for suicide risk. 

Beyond screening, there is significantly less uptake of the additional four best practices. In fact, only 8% of Joint Commission-accredited hospitals meet full criteria for implementing all recommended suicide prevention activities at the time of patient discharge. Separately, 28% of hospitals developed a plan for lethal means counseling, 30% made follow-up contact, 37% provided warm handoffs to outpatient care, and 61% implemented formal safety planning. However, only about 19% of hospitals that completed safety planning included all eight recommended components, which involve processes such as documenting warning signs and coping strategies, identifying people and settings that can provide distraction from suicidal thoughts or offer help during a crisis, and developing a plan for lethal means safety.⁵

Although hospitals are not engaging in suicide prevention practices broadly, individual hospitals and health systems are acting as leading examples for risk reduction. Intermountain Health, a nonprofit system with 34 hospitals and 400 clinics across six states in the Mountain West, is one such example. Through its participation in the Suicide Care Collaborative Improvement and Innovation Network, a collaborative learning initiative led by the Zero Suicide Institute in partnership with The Pew Charitable Trusts, Intermountain Health worked to strengthen suicide prevention training for staff and improve care transitions for discharged patients. The system had already integrated suicide risk screening and assessment into routine behavioral health care and hired behavioral health navigators to follow up with patients within three days of discharge. System data show that between 58% and 66% of patients in behavioral health settings are now seen for a follow-up visit within seven days of discharge, well above the national average of 27 to 48%. Through the initiative, Intermountain Health also improved collaborative safety planning, setting a goal that 60% of patients at risk for suicide would complete a safety plan before discharge and meeting or exceeding that target in eight of ten months tracked. Intermountain Health does this in an environment where many of these services are not reimbursable and roughly 30% of patients at the implementation site are on Medicaid. Yet health system leaders have remained committed because they believe this care is the right thing to do and that it makes a meaningful difference for patients and their families.

Economic considerations

While there is evidence showing that these practices are likely to be cost-effective from a societal perspective—meaning that the broader economic and social benefits, including those that accrue to parties beyond the patient and provider, are achieved at a cost that compares well to other uses of funds—they may add to a particular hospital’s costs. These societal benefits include, for example, reduced emergency visits to other hospitals and clinics, along with fewer deaths by suicide and the disruptive impacts such deaths have on families and communities. In some cases, for EDs, some of these practices may be reimbursable. That is less likely for inpatient units, which are typically paid via a bundled case rate (a fixed payment for the entire inpatient stay) or a bundled day rate (a fixed payment per day of care). 

The findings concerning hospitals’ fidelity to the four recommended practices represent a classic misalignment of incentives: the entity that incurs costs (in this case, the hospital) is not always the one that captures the financial benefits (which often accrue to insurers or society at large through avoided future utilization and mortality).  

Under existing payment arrangements and legal processes, there is a complex array of incentives that are likely to affect hospital decisions regarding which practices to adopt. Here we briefly summarize some of those incentives. 

Risk-sharing between hospitals and insurers

First, there are downstream costs that are frequently incurred by a hospital. Even under bundled payments, hospitals often bear some financial risk, such as readmissions and repeat ED visits. That is, higher-quality safety planning and follow-up can reduce short-term revisits. Some payment systems penalize readmissions within a 30-day window. In addition, effective discharge planning can prevent complications or crises that prolong inpatient stays. If a significant share of hospital revenues stems from risk-based contracts where the hospital shares responsibility with payers for the cost and quality of care, hospitals do internalize some of the savings from reduced future utilization. Therefore, while the full societal benefit isn’t captured, a slice of it can be, especially as payment models shift toward risk-sharing.

Risk mitigation investment

A second class of incentives is less direct: those related to liability and risk management. Suicide after discharge is a major source of malpractice claims. Demonstrating use of evidence-based practices, such as the four recommended practices, can reduce legal exposure or strengthen defenses. Hospitals often view these practices as a form of risk mitigation investment, even if they are not directly reimbursed.

Reputational risk management

A “softer” set of incentives concerns reputation and community expectations. Hospitals, especially nonprofits, have incentives tied to community benefit obligations (although these are weakly enforced in most states), brand trust and public perception, and relationships with local providers (physicians) and stakeholders. Preventing highly visible adverse events like suicide aligns strongly with these pressures. However, this reputational incentive is blunted by low public awareness: it is not widely known how often suicide follows a recent hospital encounter, nor that most hospitals have yet to implement the practices shown to reduce that risk. Greater public visibility into how much more hospitals could do could strengthen reputational pressure to adopt these practices.

Incentives to increase the adoption of best practices in suicide care

Two of the practices with the weakest fidelity to recommendations share a key feature: most of their impact occurs after the patient leaves the hospital. Warm handoffs (37% fidelity) mainly reduce missed outpatient follow-up visits and improve continuity of care, while follow-up contacts (30% fidelity) reduce suicide risk and downstream crises over extended periods (days to weeks). The credit and benefits generated from those activities often accrue to outpatient providers or insurers, not the discharging hospital. They also show up as avoided events (ED revisits, admissions, deaths), which are probabilistic and delayed, and are less visible in hospital performance metrics compared to, say, length of stay or in-hospital complications. So relative to in-hospital practices (like safety planning), the hospital captures a smaller share of the return. 

These practices are deceptively simple but operationally difficult to execute. Warm handoffs require several steps, such as (a) real-time scheduling with and availability of outpatient providers; (b) maintaining referral networks; and (c) staff time for coordination (often social workers or care managers). Likewise, follow-up contacts involve (i) staffing (for calls, texts, tracking); (ii) systems for outreach and documentation; and (iii) protocols for escalation if risk is detected. These are ongoing, labor-intensive processes, not one-time clinical actions. And importantly, they don’t typically fit neatly into existing hospital workflows, which are optimized for episodic, facility-based care. 

Lethal means counseling (28% fidelity) is infrequently adopted by hospitals due to a combination of system-level pressures, provider discomfort, lack of training, and cultural and perceived legal barriers. This low fidelity occurs even though hospitals that do implement routine lethal means counseling are associated with significantly lower suicide rates following discharge, highlighting that low adoption represents a missed opportunity for suicide prevention. 

Even when benefits do accrue to the hospital (such as via fewer readmissions), attribution is difficult to establish. For example, if a patient doesn’t return, was that the result of the follow-up outpatient visit or family support? If the patient does return, the intervention may look ineffective even if it reduced severity. That makes these practices a harder internal “sell” compared to interventions with clear, attributable outcomes inside the hospital. 

As noted earlier, inpatient stays are often paid via bundled case rates or per diem, with no explicit payment for post-discharge work. EDs may sometimes bill for certain services, but follow-up outreach is rarely reimbursed. Underinvestment in this type of care is very common, as the benefits are diffused, uncertain, and often external. 

To increase adoption of the four suicide prevention best practices, changes in incentives or payments would need to shift at least one of the following margins: 

• Increase captured benefits (e.g., through shared savings or bundled episodes that include the post-discharge period). 

• Reduce costs (e.g., through centralized follow-up services or digital outreach), though this is less likely. 

• Improve attribution and measurement (e.g., through quality metrics tied to transitions out of the hospital setting).  

• Extend accountability (e.g., through episode-based or population-based payment).  

Absent such changes, hospitals will rationally continue to limit investments in these practices, particularly those that occur “beyond the four walls” of the hospital itself.

Towards implementation of recommended practices via an incentives strategy: Choosing an approach

As detailed in previous sections, despite their effectiveness and likely societal cost-effectiveness, adoption of the four practices remains incomplete. In this section, we outline key considerations and propose a framework for an accountability and payment system designed to increase adoption and fidelity of these practices while addressing heterogeneity across hospitals and avoiding inefficient or inequitable payments. Throughout, payment incentives are tied to the documented implementation and delivery of the recommended practices, not to downstream clinical outcomes such as suicide rates or readmissions.

Challenges

Developing a payment and accountability system aimed at hospital suicide prevention practices requires grappling with five structural challenges. 

The first involves heterogeneity in hospital capacity. Hospitals vary substantially in financial resources, staffing, behavioral health integration, and access to outpatient networks. Resource-constrained hospitals, particularly safety-net and rural providers, face higher fixed and marginal costs of implementing care coordination practices. Rural hospitals face a distinct combination of constraints: thin or negative operating margins, chronic behavioral health workforce shortages that make hiring dedicated staff for safety planning and follow-up especially difficult, and sparse outpatient networks that can make warm handoffs difficult even when hospital-side protocols exist. As a result, recent research shows that rural hospitals and nonprofits are substantially less likely to engage in post-discharge suicide prevention services. 

The second challenge relates to an earlier observation about misaligned incentives such that hospitals bear the costs of implementation, while many benefits, such as reduced mortality and downstream utilization, accrue to payers or society. This misalignment is most pronounced for warm handoffs and follow-up contacts.  

Third, hospitals begin at widely different levels of adoption. Therefore, a uniform performance target risks over-rewarding high performers without inducing change and offering too weak an incentive to “low performers,” who are especially relevant for reaching many key populations, including those in rural areas or with complex social needs. 

Fourth, pure improvement-based incentives may reward small, clinically insignificant gains among very low performers, while failing to ensure meaningful standards of care.  

Finally, not all four practices are equal in their difficulty to implement. Practices that extend beyond the hospital (“beyond-the-walls”) are more costly, operationally complex, and depend on external provider capacity. 

These challenges collectively suggest that no single payment approach will be sufficient. Any viable model must be flexible enough to account for differences in hospital capacity, baseline performance, and practice complexity. We consider two broad approaches before proposing a hybrid approach. 

Payment approaches

Here we consider two main approaches to payment: 1) a graduated bonus (upside-only) model, in which hospitals earn increasingly large bonus payments as they improve performance, but face no financial consequence for failing to do so; and 2) a two-sided risk model, in which hospitals share financial accountability in both directions—they earn bonuses for meeting or exceeding performance targets but face penalties when performance falls short of those targets or fails to sufficiently improve. 

A one-sided (bonus-only) graduated bonus has the advantage that it encourages participation, minimizes financial risk, and is useful during the early stages of implementing an accountability system, as it reduces the risks of a sluggish start and the uncertainties of a new set of policies. This type of bonus, however, also carries some notable limitations. Most significant is that, because there is no penalty for not improving performance, “moving the needle” may require substantial payments to motivate changes in hospital practices. In addition, there is the risk of paying for marginal (low-value) improvements. 

The two-sided risk model creates stronger incentives for hospitals to implement the recommended practices. It puts penalties in place for persistent underperformance. However, such a payment approach has been shown, in other contexts, to deter participation by lower-capacity hospitals, which risks exacerbating existing disparities across hospitals and the communities they serve. 

These considerations suggest that a phased approach—beginning with upside-only incentives and transitioning to two-sided risk for capable hospitals—may be most effective.

A mixed payment approach

A hybrid approach combining improvement-based rewards, attainment thresholds, and practice-specific weighting may be the most effective approach. Under this model, each of the four practices is measured and rewarded separately rather than through a single composite score. The size of the reward for each practice would be set according to its relative difficulty and operational complexity. This could mean assigning a higher payment for improvements in warm handoffs and follow-up contacts, encouraging hospitals not to disproportionately focus on lower-cost, within-hospital practices like safety planning. That means using dual metrics that at once pay for improvement to engage and encourage behavior change in hospitals that have been most reluctant to implement the recommended practices, while also requiring a minimum attainment threshold to realize larger bonuses. The attainment standard seeks to ensure clinically meaningful performance. This payment structure implies that small improvements at very low baseline levels receive limited payments, and larger rewards are tied to reaching higher absolute performance levels. 

Thus, the use of a tiered or graduated payment structure recognizes early modest improvements, but compensates them modestly and accordingly. Payments increase as performance approaches target levels. This serves to reduce incentives for minimal compliance while preserving motivation for progress.

Addressing heterogeneity

We consider addressing heterogeneity across hospitals and with respect to the complexity of implementation across the recommended practices. Hospital heterogeneity can be addressed through 1) peer group benchmarking or 2) adjustment factors that recognize differences in hospital capacity and resources. Hospital peer group benchmarking involves creating groups of hospitals with similar attributes, for example, teaching status, safety-net designation, or rural location. Performance expectations and assessment of improvement would be adjusted to reflect those circumstances, so the performance standard would focus on performance relative to one’s peer institutions. Expectations for hospitals would then reflect structural constraints. Adjustment factors serve a similar purpose, providing higher payments or lower attainment thresholds for resource-constrained hospitals. 

Rural hospitals warrant particular attention in program design, as workforce shortages, sparse outpatient networks, and thin margins limit their ability to build capacity. Centralized infrastructure—such as state-run follow-up programs like Colorado’s Follow-Up Project, offered at no cost to hospital sites, or telehealth-based warm handoffs—can substitute for capacity rural hospitals cannot sustain internally, and quality metrics should credit participation in such shared services. Upfront infrastructure payments (or grants) would also matter more here than retrospective bonuses, given the hospitals’ limited ability to front training and staffing costs. Finally, rural hospitals’ heavy reliance on Medicare and Medicaid revenue makes CMS uniquely positioned, through state Medicaid programs and rural-specific payment mechanisms, both to expect these practices and to finance the capacity to deliver them. 

A caution

Financial incentives alone are unlikely to achieve widespread adoption, particularly for the more resource-intensive practices. One realistic complementary strategy could be establishing a source for technical assistance and standardizing protocols to guide implementation steps. 

Therefore, a staged approach might follow three phases:

Phase 1: Capacity building and engagement 

• Upside-only incentives 

• Emphasis on improvement 

• Infrastructure support 

Phase 2: Performance consolidation 

• Introduction of attainment thresholds 

• Increased weighting on high-impact practices 

Phase 3: Accountability and sustainability 

• Selective transition to two-sided risk for high-capacity hospitals 

• Integration into broader value-based payment models 

A blended model—combining improvement-based rewards, attainment thresholds, nonlinear payments, and targeted incentives for harder-to-implement practices—offers the most promising path forward. Absent such design features, payment systems risk either overpaying for marginal improvements or failing to induce meaningful changes in hospital behavior, particularly for practices whose impact extends beyond the hospital walls and that are most critical for reducing suicide risk.

Quality metrics

Overview, framework, and state lessons

Effective payment incentives for hospital quality improvement must be tied to measurable, operationalizable metrics. This section details quality metrics corresponding to the four recommended discharge practices described in this paper—safety planning, lethal means counseling, warm handoffs, and follow-up contact—with Colorado’s Hospital Quality Incentive Program (HQIP) serving as one key example for how states can structure meaningful, tiered payment incentives that produce quality improvements while accounting for heterogeneity in hospital capacity and readiness. 

Before detailing the metrics for each practice, it is important to situate them relative to existing regulatory requirements. The Joint Commission publishes NPGs, which hospitals are required to abide by to keep their accreditation. The current goals establish a foundational floor for hospitals: screening, validated assessment, documented risk level, written policies, and discharge planning. These function as preconditions, meaning hospitals should already be meeting this baseline before quality metrics tied to payment incentives come into effect. The four practices described here represent the next layer above that baseline, operationalizing what the NPG framework gestures toward, such as “plan counseling and follow-up care at discharge,” but does not prescribe in any specific or measurable way. As noted throughout this paper, these practices are strongly recommended by suicide prevention experts but are not explicitly required by the Joint Commission. 

Colorado’s HQIP, administered through the state Medicaid agency, illustrates how quality metrics can be structured at the state level. The Colorado Hospital Affordability and Sustainability Enterprise (CHASE) established HQIP payments funded through health care affordability and sustainability fees, with a subcommittee responsible for recommending quality incentive approaches. Within this program, the Zero Suicide framework was introduced to the Patient Safety measure group in 2021, using a tiered deliverable structure that progresses from leadership and planning, through training and clinical engagement and treatment, and ultimately to transition and improvement. 

Colorado’s HQIP is used throughout this section as the primary illustrative model because it is, as of this writing, the only state-level program that publicly documents specific, tiered quality metrics for suicide prevention tied directly to hospital incentive payments. Other states have pursued related quality improvement payment approaches (and their experiences are instructive), but none has published a comparable set of operationalized, payment-linked deliverables for hospitals. 

Other states pursuing suicide prevention approaches include Vermont, Michigan, and Oregon. Vermont’s Program for Quality in Health Care has led statewide quality improvement initiatives focused on suicide care in emergency departments, and Vermont’s legislature passed Act 56 in 2023, directing the development of a model protocol for health care facilities covering general standards of safety planning, lethal means counseling, care transitions, and follow-up contact, though these standards are not publicly tied to payment. Michigan’s MI Mind, a Collaborative Quality Initiative led by Henry Ford Health with Blue Cross Blue Shield of Michigan support, uses commercial payer infrastructure to tie provider reimbursement to Zero Suicide implementation through a scorecard of participation and quality improvement requirements, though the specific clinical performance metrics are not publicly documented. Oregon has also developed a structured self-assessment instrument for Zero Suicide fidelity, but it is voluntary and not payment-linked. These examples reflect a landscape in which payment-linked quality improvement for suicide prevention is limited—making Colorado’s model the most fully operationalized reference point available.  

Colorado’s experience illustrates the potential for quality incentive payments to address exactly the implementation barrier described earlier. Hospitals that successfully completed Zero Suicide deliverables through the HQIP program earned substantial quality incentive payments, in some cases sufficient to fund the hiring of behavioral health specialists dedicated to care navigation and follow-up.⁶ The incentive payment, in other words, made the staffing investment financially viable. The positive cycle this creates—earning incentives by demonstrating commitment to evidence-based practices, then using those funds to build the infrastructure for those practices—is central to the appeal of the Colorado model.  

Quality metrics for the four practices below could be structured similarly: building from foundational organizational commitment toward increasingly operationalized clinical practices. Hospitals can earn points and associated payment incentives for successfully completing deliverables at each level and must complete earlier levels to be eligible for higher-level points. Crucially, hospitals still earn something for meeting foundational expectations even if they have not yet reached more advanced deliverables, which sustains participation among hospitals at earlier stages of implementation. This tiered structure, where meaningful money is nonetheless on the table, is central to the model’s effectiveness.

Safety planning

Safety planning is one of the most commonly reported discharge practices among hospitals—61.3% of surveyed hospitals report some implementation—yet very few implement it with fidelity to all key components; of those, only around 19.3% documented all required elements of a complete safety plan. This gap between partial and full implementation makes safety planning a strong candidate for a quality metric that assesses not merely whether a plan was completed, but whether it was completed with fidelity to the evidence base. 

Research has identified eight key components of a complete safety plan:  

1. Promoting shared understanding of its purpose 
2. Identifying and documenting warning signs
3. Documenting internal coping strategies 
4. Identifying people and social settings that provide distraction 
5. Identifying personal contacts who can provide help 
6. Documenting professional services to contact during a crisis 
7. Developing a plan for lethal means safety 
8. Explaining how to best use the safety plan 

A meaningful quality metric would assess fidelity to all eight components, not just whether a plan was created, but whether each element was addressed, documented, and provided to the patient. 

Timing is an equally critical dimension of quality care. One key quality metric is whether a safety plan was completed within an appropriate amount of time, such as within 24 hours after a patient screened positive for suicide risk, and whether it was reviewed and updated as part of discharge planning before the patient leaves the facility. Existing fidelity instruments, such as Oregon’s Zero Suicide Implementation Assessment, address safety planning at the point of risk identification and modification at subsequent visits but stop short of explicitly requiring it as a discrete step at discharge. 

One practical path toward measuring fidelity and designing trackable quality measures is embedding the safety planning tool directly into the electronic health record (EHR). When EHR fields correspond to each component of the safety plan (i.e., a structured checklist aligned with Stanley-Brown Safety Planning Intervention components), it becomes possible to track completion at the field level and identify which components were addressed for which patients. Health systems implementing Zero Suicide have found it challenging to develop efficient ways to track the content and completeness of safety plans; some have created specific EHR sections where safety plans are recorded in detail, while others develop plans on paper and upload images to the record, which makes completeness much harder to assess. A study of safety planning measurement across six health systems implementing Zero Suicide found that structured EHR templates with discrete, non-customizable fields—a checklist-like approach—are more accurate and less costly for quality tracking than relying on narrative clinical notes, which require complex and resource-intensive natural language processing tools to assess at any scale. However, only three of the six systems studied implemented these tools. Hospitals will inevitably approach this differently depending on their existing EHR infrastructure, workforce capacity, and stage of Zero Suicide implementation, but that variation is to be expected and quality metrics should reflect this range.  

There are also practical constraints worth acknowledging. Some providers have reported that time pressure makes thorough safety planning difficult, particularly when suicidal ideation is not disclosed until late in a clinical encounter. Some organizations have responded by permitting an abbreviated safety plan when a patient presents as lower risk, with a complete plan developed at the next available session. This kind of adaptive protocol—accommodating clinical realities while maintaining a clear expectation of eventual completeness—may need to be reflected in how quality metrics are designed, so that pragmatic flexibility does not simply become a loophole for incomplete practice.  

Finally, an important but often overlooked dimension of safety planning implementation is workforce scope. With appropriate training and supervision, safety planning does not have to be performed exclusively by licensed behavioral health clinicians, and in a performance-based payment model, this flexibility becomes more achievable.⁷ When hospital facilities or programs are paid based on quality outcomes of the organization rather than traditional individual service billing, the traditional constraints that tie specific procedures to specific credentialed staff are less determinative.⁸ Expanding the pool of qualified staff who can conduct safety planning may be a structural factor worth addressing, but the path to doing so is more straightforward when payment is attached to a hospitals’ overall quality performance rather than the billing privileges of any individual provider. 

With these constraints in mind, for quality metric purposes, we recommend a tiered structure anchored to both the timing and the content of safety planning. At a minimum, hospitals should be required to document that a safety plan was completed during the same encounter in which a patient screened positive for suicide risk, and that an updated plan is reviewed and provided to the patient as part of discharge or care transition planning before the patient leaves the facility. At each of these points, the plan should be assessed for completeness across all eight components outlined previously, shared with the patient, and, where consent permits, with family members or caring contacts, and modified as warranted. A checklist-based EHR template mapped to the eight Stanley-Brown Safety Planning Intervention components is a recommended standard because it produces the field-level, auditable data that fidelity tracking requires. For hospitals earlier in implementation, a paper-based plan with same-day transcription into structured EHR fields is an acceptable interim approach. The highest tier of payment should be given to hospitals that demonstrate both complete and adequately documented safety plans administered at both key time points—first engagement and discharge—for patients at varied levels of suicide risk.

Lethal means counseling

Lethal means counseling has long been understood as a component of comprehensive safety planning, but the evidence supports treating it as a separately measured and incentivized practice. When bundled inside a broader safety planning requirement, lethal means counseling is the element most frequently omitted. Only approximately a quarter of hospital safety plans include a lethal means component, and rates of documented lethal means assessment in practice are lower still (only 18% of ED patients with a positive screen were assessed; of those, only 8% had a documented action plan; and among patients discharged home, 85% had no documentation at all). 

A discrete quality metric for lethal means counseling could include: documentation that access to lethal means—including firearms, the most common method of suicide death, along with other methods—was assessed; documentation of counseling on safe storage practices; and documentation of an action plan for temporarily limiting access following discharge. The metric should apply not only to the inpatient encounter but also extend explicitly to discharge planning, confirming that a concrete plan was in place and documented before the patient left the facility. 

One promising mechanism for improving documentation is the integration of lethal means assessment and counseling prompts directly into the EHR, a practice that some health systems have adopted as part of Zero Suicide implementation. Where such prompts are in place, lethal means service delivery improves, suggesting that structural supports—not just clinician awareness—drive fidelity. EHR integration of lethal means prompts could therefore serve as both a facilitative best practice and a prerequisite deliverable within a tiered metric structure, analogous to the role it plays in safety planning more broadly. 

We recommend that lethal means counseling be measured as a discrete quality metric rather than treated as only a component of safety planning, given the consistent evidence that it is an element omitted when bundled inside broader requirements. At a minimum, the metric should require documentation that access to lethal means—including firearms and toxic substances—was assessed within each safety planning occurrence; furthermore, that counseling on safe storage practices occurred, and that a concrete action plan for limiting access was in place and documented before the patient left the facility. Consistent with the timing standards recommended for safety planning, lethal means counseling should be completed as part of the initial clinical encounter following a positive risk screen and explicitly revisited as part of discharge planning. Staff training and use of an evidence-based lethal means counseling approach, such as Counseling on Access to Lethal Means (CALM), should be a prerequisite for meeting the foundational tier, with documentation of the percentage of at-risk patients receiving a complete lethal means assessment and discharge action plan required for higher-tier payment.

Warm handoff

Providing a phone number or including a referral in discharge paperwork does not constitute a warm handoff, and quality metrics could make this distinction clearer. The post-discharge period is among the highest-risk windows for suicide, and research demonstrates that patients who receive connections to supportive follow-up care in the days immediately following discharge are less likely to die by suicide. 

The central challenge of designing a warm handoff metric is that its success depends in part on factors outside the hospital’s direct control, such as the availability and responsiveness of outpatient providers. Metrics should therefore focus on what the hospital can directly control: whether discharge planning began early in the inpatient stay; whether the hospital has formal collaborative protocols with outpatient partners, such as memoranda of understanding; whether a follow-up appointment was scheduled prior to discharge; and whether a clinician-to-clinician handoff was initiated that attempted to connect the patient with the outpatient provider, most often by telehealth, before discharge. These are all actions that can be taken within the hospital, even if the outpatient partner’s responsiveness is not fully within its control. 

Colorado’s HQIP program provides a useful approach: the warm handoff deliverable requires hospitals to complete a self-assessment checklist from the National Action Alliance for Suicide Prevention and submit a work plan for improving collaborative protocols with outpatient organizations. The checklist describes a range of practices, from basic administrative preparation to active care coordination, rated on a scale reflecting proximity to best practice. This range-based structure is valuable because it acknowledges variability in hospital capacity and existing outpatient relationships, while still pushing hospitals to identify gaps and plan improvements. 

The checklist addresses both administrative and clinical dimensions. On the administrative side, it covers whether the hospital has formal agreements with partner organizations, whether it regularly meets with those partners to share metrics, and whether data-sharing protocols are in place. On the clinical side, it covers when discharge planning begins during a stay, whether family members and peer specialists are included, whether a follow-up appointment is scheduled before discharge, and whether the patient is introduced directly to the outpatient provider, either in person or via telehealth, before leaving. Any of these elements could be weighted in a quality metric, with financial incentives scaled to how many are achieved and how close to best practice the hospital’s processes are. 

Some states and systems have explored a model in which a centralized state program or contracted entity facilitates the handoff on behalf of hospitals, reducing the real-time coordination burden on hospital staff.⁹ In such a model, a hospital clinician contacts the program, hands the phone to the patient, and the program takes over—facilitating the introduction and, in some cases, subsequent follow-up contacts. While this model can ease implementation, it is not a true clinical handoff to an outpatient provider; the quality of these programs warrants its own monitoring.¹⁰

We recommend that quality metrics for warm handoffs focus on what the hospital directly controls, structured in a tiered progression from foundational administrative readiness toward active clinical handoff protocols and best practices. At a minimum, hospitals should be required to document that written policies and protocols for warm handoffs exist: specifying who is responsible for initiating the handoff, how outpatient follow-up is communicated to the patient, and how the process is documented; that either formal collaborative agreements with outpatient behavioral health partners are in place or the hospital has designated in-house staff with the capacity to facilitate these care transitions directly; and that some form of outpatient connection is made before or at the time of discharge, even if the timing and modality vary. At the next tier, hospitals should demonstrate that a follow-up appointment was scheduled within 24 to 72 hours of discharge, with the appointment itself occurring no later than seven days after discharge. 

At the best-practice tier, hospitals should document that direct contact between the patient and the receiving outpatient provider occurred before discharge, via telehealth or in person. This clinician-to-clinician handoff is the evidence-based standard. Centralized intermediary programs, in which a hospital clinician connects the patient to a contracted program rather than directly to the receiving provider, can serve as a temporary option for hospitals without established outpatient partnerships. However, they should be reimbursed at a lower tier and treated as a bridge, not an endpoint. In all cases, hospitals should document both the scheduled appointment and the handoff before the patient leaves the facility. A phone number or referral listed only in discharge paperwork should not count as meeting this standard.

Follow-up contact

Research has consistently found that supportive contacts following psychiatric hospitalizations or ED visits can help reduce suicide, though the evidence base has not yet identified the most effective approach in terms of timing, frequency, or content. The National Action Alliance for Suicide Prevention recommends that the first follow-up contact post-discharge occur within 24 to 48 hours, but the optimal number and frequency of subsequent contacts, particularly in the period bridging discharge and a patient’s first outpatient appointment, are less clearly specified by the evidence. Follow-up contacts have taken a range of forms, from a reminder of an upcoming appointment location and time, to a structured reassessment of suicide risk, coping skills, and collaborative safety planning. This variability reflects both the flexibility of the intervention and the challenge health systems face in defining what constitutes adequate and effective follow-up, and how to monitor it. 

Quality metrics for follow-up contact should address several dimensions: timing (how soon after discharge is the first contact made); modality (phone call, telehealth, text, or other caring contact); consistency (how many contacts are made and whether they are sustained until the patient has successfully connected with outpatient care); and personnel (whether contacts are made by someone the patient has an existing relationship with, to support rapport). A tiered metric structure could be built from a minimum expectation—one follow-up contact within 24 to 48 hours of discharge, consistent with existing guidance—toward a best practice of multiple contacts sustained through successful outpatient engagement. 

Colorado’s HQIP program offers two approaches to design this quality metric. The primary deliverable requires hospitals to demonstrate participation in the Colorado Follow-Up Project, a telephonic caring contact program serving patients discharged from EDs or inpatient psychiatric units following a mental health crisis or overdose, offered at no cost to hospital sites or enrolled patients. Participation in a program of this kind simplifies tracking—the hospital’s obligation is to connect the patient to the program and confirm enrollment each month. The alternative deliverable, for hospitals not participating in the Follow-Up Project, requires a written policy and work plan for follow-up contacts within three days of discharge for patients who screened positive, including identification of responsible staff and a tracking system, along with annual reporting on the number and percentage of patients who received such contacts. 

A related question worth addressing in metric design is the format of follow-up contact. Phone calls remain the most common modality for follow-up care. Telehealth, text-based check-ins, and written caring contacts have expanded in use, particularly following the COVID-19 pandemic, and offer potential advantages in reach and flexibility. Quality metrics should allow for flexibility in format while specifying that any contact must be a genuine, interactive, caring contact—not an automated message—and that the content should include, at minimum, an assessment of how the patient is doing and reinforcement of next steps for outpatient connection. 

We recommend that follow-up contact quality metrics be structured in a tiered progression that reflects both the operational realities hospitals face and the evidence on what constitutes meaningful post-discharge support. At a minimum, hospitals should be required to document a written policy specifying who is responsible for follow-up contacts, what the content of those contacts should include, and how they are tracked, with at least one contact made within 24 to 48 hours of discharge for all patients who screened positive. 

At the next tier, contacts should be risk-stratified and sustained until the patient has attended their first outpatient appointment: for example, high-risk patients could receive contact within 24 hours of discharge, again within 72 hours, and at least weekly until outpatient engagement is confirmed; moderate-risk patients within 48 hours and at least once more before their first appointment; and lower-risk patients at least once within 48 hours confirming appointment details and current status. Each contact should include, at a minimum, a brief risk assessment, safety plan review, appointment reminders, and active support for navigating barriers to attendance, not just a simple check-in call. Protocols should address no-shows with documented outreach within 24 hours of a missed appointment.  

At the best practice tier, which we recommend as the target standard, hospitals should demonstrate that contacts are conducted by a clinician or staff member trained in safety planning; that follow-up is sustained until the patient has verifiably attended their first outpatient appointment; that the capacity for in-person or virtual home or community visits exists for high-risk or hard-to-reach patients; and that the organization has established approaches—such as peer supports, peer-run crisis respite, or drop-in appointments—to address patients who are difficult to reach by phone.¹¹ Quality metrics should allow for flexibility in format while specifying that any contact must be genuine and interactive.

Implementation considerations

Across all four practices, health systems implementing Zero Suicide have found that robust tracking and continuous quality improvement are essential to sustaining fidelity over time. Organizations that have implemented Zero Suicide successfully tend to conduct regular audits and accountability checks, and to hold post-incident reviews when needed. EHR modifications, such as automatic reminders to complete screeners or follow-up assessments, have helped address gaps identified through monitoring. 

Implementation also needs to account for the fact that a hospital’s obligations under these four practices look different depending on whether a patient is discharged from the ED without admission or discharged after an inpatient stay. Inpatient care allows more time to identify an outpatient provider, initiate a warm handoff, and confirm a follow-up appointment before the patient leaves. ED-based care offers a narrower window: safety planning and lethal means counseling can generally still occur before discharge, but completing a clinician-to-clinician handoff or securing a confirmed appointment in that same visit may not be feasible—particularly for patients without an existing outpatient relationship—and hospitals should not be incentivized to board or admit patients solely to complete a practice that could otherwise be finished shortly after ED discharge. Tracking and quality improvement infrastructure should reflect this distinction as well: EDs and inpatient units often rely on different staff, workflows, and systems to carry out these practices, and hospitals may need to build separate tracking and follow-through processes for each setting rather than assuming a single infrastructure investment will serve both. 

Outcome tracking presents additional challenges. Many organizations have sought to track suicide-related outcomes such as hospitalizations, emergency department visits, and deaths, but capturing a complete picture is difficult when many outcomes occur outside the health system. Integrated health systems have more capacity to track these outcomes across settings, while outpatient organizations face greater limitations. Some organizations have addressed this by linking patient records to state death records, partnering with community surveillance efforts, or working with their state hospital association to gather better data on suicide-related hospitalizations. Stakeholder discussions have raised the aspiration of tracking survival from suicide attempts the way cancer and cardiac events are tracked, which is done longitudinally and through linkage with state mortality data, as a longer-term outcome metric that would complement process-level quality measures.¹²

For quality metrics tied to payment incentives, process measures are likely to be the most operationalizable in the near term: whether a safety plan was completed with fidelity, whether lethal means were assessed and a discharge plan documented, whether a warm handoff was initiated, and whether follow-up contact occurred within a specified window. Process metric examples from Zero Suicide implementation research include: outpatient follow-up appointment scheduled within 48 hours of discharge; patients who attended a follow-up appointment within 48 hours of discharge; and follow-up calls completed with a patient. These process quality measures provide a practical starting point, with the expectation that data infrastructure and longitudinal outcome tracking will develop over time as hospitals build capacity and experience. 

The quality metrics proposed here are intended as illustrative examples of what a payer could require. However, hospitals vary considerably in their baseline capacity, existing outpatient partnerships, EHR infrastructure, and behavioral health workforce. A well-designed quality metric program should reflect this by setting floors that are achievable for most hospitals in an initial program year, using tiered structures that reward progress toward best practice, providing hospitals with technical assistance and implementation tools, and building in annual reporting requirements that generate data to inform future refinements. The incentive payment itself must be substantive enough to justify the administrative investment of meeting deliverables and meaningful enough to fund at least incremental capacity-building. Quality metrics tied to negligible payments are unlikely to change practice at scale. As the Colorado HQIP model demonstrates, when the incentive is real, voluntary participation can generate meaningful system-wide change.

Payment design

Given what we know about the hospital setting, options for payment approaches, and quality metrics that are relevant for creating incentives for hospitals to engage in suicide prevention best practices as discussed above, the goal of this section is to provide specific details on what payment design might look like. In particular, payment should reflect an approach incorporating two-sided risk and bonuses based on cost and quality targets that incentivize hospital participation in all four best practices, including those that have outcomes “beyond the four walls.”

Lessons from the Center for Medicare and Medicaid Innovation

Several payment models from the Center for Medicare and Medicaid Innovation (CMMI) incorporating two-sided risk and quality bonuses have been successful in recent years. Two examples, the Comprehensive Care for Joint Replacement (CJR) Model and the Enhancing Oncology Model (EOM), are particularly relevant to learn from, as both involve care coordination and engaging other providers outside of the initial hospital or physician practice. Payment models such as these also have applications in the mental and behavioral health setting. A recent paper highlighted how lessons from CMMI, particularly the CJR Model, can be applied to improve Medicare payments to opioid treatment programs (OTPs).  

The EOM involves participating physician groups taking on two-sided risk for quality and spending during six-month episodes of care for Medicare patients with a set of high-risk cancer types, in an effort to improve quality and care coordination. In this arrangement, practices receive performance-based payment tied to financial accountability, as well as a payment per beneficiary per month intended to support activities such as around-the-clock access to clinicians, patient navigation, comprehensive care planning, screening for health-related social needs, and collection and monitoring of electronic patient-reported outcomes. To provide these services, the physician practices coordinate care with other specialists, primary care providers, community organizations, social workers, and psychologists. At the end of the six-month period, participating physician groups keep some savings if their spending falls below the target amount while quality metrics are met. 

The CJR Model similarly encourages quality improvement and care coordination. Through the CJR Model, participating hospitals receive a retrospective bundled payment for taking on additional care coordination responsibilities for patients undergoing hip, knee, or ankle replacements—specifically, coordinating with physical therapy and other rehabilitation care. The bundled payment is meant to incentivize the avoidance of expensive and harmful events and is therefore based on quality measures such as the complication rate following surgery, information from a consumer assessment survey, and patient-reported outcomes. In efforts to improve the effectiveness of care coordination, CMS provides participating hospitals with relevant spending and utilization data, waives certain requirements for care delivery, and facilitates the sharing of best practices between hospitals.  

Although the federal government administers CMMI models, states are also engaging in innovative payment strategies to incentivize high-quality health care, including suicide prevention services. For example, Colorado’s HQIP and Hospital Transformation Program (HTP) have both been leveraged within Colorado hospital systems to implement the Zero Suicide framework and deliver high-quality suicide prevention care, as well as to build out a behavioral health workforce capable of supporting high-fidelity adherence to these best practices.¹³ As touched on earlier in the paper, HQIP and HTP both provide incentive payments through Medicaid tied to hospital performance on a variety of quality outcomes. HQIP uses a tiered funding approach aligned with the elements of the Zero Suicide framework. Certain core elements must be met to receive any payment, which is meant to incentivize fidelity to the entire host of Zero Suicide components and, in turn, to provide the highest quality of care. 

Costs of implementing a bundled payment

Although data on the costs to hospitals of completing all suicide prevention best practices are sparse, various estimates for a set of services can provide insight into a reasonable amount for a bundled payment that might be included in an innovative payment model for hospitals. Evidence suggests that across different populations, universal screening costs a few dollars per person who is screened in the ED.¹⁴ This order of magnitude is reflected in the Medicare reimbursement rate for administering a brief, standardized emotional or behavioral assessment (CPT code 96127), which has a national payment amount of about $5 per administration in 2026.  

Beyond implementing universal screening, one study found that providing enhanced suicide risk assessment, the provision of a self-administered safety plan, mental health treatment information, and suicide hotline resources at discharge, and follow-up contact through a series of phone calls cost about $500 per participant. Again, this cost aligns closely with aggregated national Medicare reimbursement rates in 2026 for these services. A 20-minute safety planning intervention is billable at around $40 (HCPCS code G0560), a set of four telephonic follow-up contacts per month is just under $67 (HCPCS code G0544), and transitional care management requiring communication within two days and a face-to-face visit within 7 or 14 days, depending on the complexity of medical decisionmaking ranges between $220 (CPT code 99495) and $300 (CPT code 99496).  

Further estimates indicate similar orders of magnitude for costs associated with different kinds of follow-up care and warm handoffs. The cost of follow-up via eight postcards was estimated to be $145, accounting for $10 per person for the cards, plus an additional $135 in additional outpatient treatment resulting from receiving the caring contact. Telephone outreach costs $300, including $30 for the phone calls themselves, plus an additional $270 for outpatient treatment resulting from receiving the calls. A final intervention, cognitive behavioral therapy, was estimated to cost $810 for nine sessions.  

Overall, a bundled payment might exceed the costs of delivering screening, safety planning, lethal means counseling, follow-up contact, and warm handoffs. Any bundled payment must treat startup costs as a distinct and significant consideration, separate from the ongoing marginal costs of delivering these services. For hospitals in the early stages of delivering these services, the initial investment in training clinicians and staff is substantial, not because the trainings themselves are costly, but because the clinician time diverted from billable care represents significant revenue loss. This is particularly consequential given that medical education has historically underprepared clinicians for mental and behavioral health encounters, meaning that training is not optional but foundational to any meaningful implementation effort. A bundled payment that accounts only for the ongoing delivery of services without recognizing these upfront costs will systematically disadvantage hospitals that are just beginning to build this infrastructure.¹⁵

Beyond startup costs, payment should account for the marginal cost of hospitals providing these services—which is partially reflected in the data presented above. Suicide screening, safety planning, and lethal means counseling tend to have lower costs, as these services are located within the hospital itself. Comparatively, follow-up contact and warm handoffs have larger costs, as these services involve connecting patients to care outside of the hospital or engaging with them beyond an initial event, requiring more resources. As such, a payment model might account for these differing marginal costs through the difficulty of achieving their associated quality metrics.

Example of a bundled payment strategy

Consider the following as a specific example of a possible bundled payment strategy: An incentives program for hospitals to employ suicide-prevention best practices would involve two-sided risks and quality bonuses based on containing costs while meeting quality metrics and reporting standards. A hospital would be required to report on quality metrics for screening and the four best practices—for example, rates of suicide risk screening among all patients presenting in the emergency department; fidelity to all eight required components of safety planning; documentation of lethal means counseling, including assessment of access to lethal means, safe storage practices, and action planning; hospital work plans for collaborative protocols with outpatient organizations to facilitate warm handoffs; and the frequency of follow-up calls with a patient after discharge. Hospitals would also have a spending target for engaging in screening and the four best practices, roughly on the order of $1,000 per patient per visit to the ED with a high risk of suicide. Hospital costs relative to this average, alongside reported data on quality metrics, would determine bonuses or repayments for performance. High-performing hospitals that met the cost target while also submitting high-quality data reflecting strong fidelity to best practices could receive up to a 25% bonus on the $1,000 target.¹⁶ On the other hand, hospitals that spent above the cost target, that did not submit high-quality data and did not reflect fidelity to all quality metrics, could face up to a 25% repayment on the $1,000 target. It is important to note that this incentive program does not need to be cost-neutral, meaning that all high-performing hospitals would receive bonuses regardless of whether there are enough low-performing hospitals to balance all costs of administering the program. 

A comment on the role of federal payers and health plans

Much of what has been discussed here involves changing accountability and payment arrangements in ways that promote suicide prevention actions by hospitals. The American health care system is a decentralized, pluralistic system of payers, health insurers, and providers. This fragmentation makes creating incentives that are common and consistently applied across the health care delivery system challenging. This point is reinforced by the fact that key actions involve the interaction of hospitals with other health care providers. 

However, hospitals rely critically on revenues from health plans attached to Medicare, Medicaid, and the ACA Marketplaces, which create opportunities to align accountability and payment across the U.S. hospital system. Recent estimates show that just under half of hospital revenues are accounted for by health plans and payments overseen by CMS: Medicare accounts for 25% of hospital revenues, Medicaid 19%, and ACA Marketplace plans 4%.¹⁷ The implication is that the federal government is in a position to require health plans serving Medicare, Medicaid, and the Marketplaces to meet performance targets aimed at hospital suicide prevention actions. The federal government could also include provisions in the Medicare Inpatient Prospective Payment System (IPPS) that align with the payment arrangements and metrics outlined earlier. This means that the federal government, through Medicare Advantage, state Medicaid plans, Marketplace rules, and Traditional Medicare Part A payment arrangements, influences a sufficiently large portion of hospital revenues to motivate hospital actions related to suicide prevention.

Conclusion

Overall, there is significant room for hospitals across the country to increase fidelity to suicide prevention practices for patients who come through their emergency departments and inpatient units with high risk. Alongside universal suicide risk screening, these practices include safety planning, lethal means counseling, warm handoffs to outpatient care, and follow-up contact. While engaging in these practices promotes the highest degree of patient safety, hospitals face a variety of constraints that prevent them from doing so. As such, payers can play a role in incentivizing hospital action through innovative payment mechanisms tied to performance on quality metrics and cost. 

We suggest that payers develop a two-sided risk model with quality bonuses, similar to those existing in Medicare through CMMI and state Medicaid programs, such as Colorado’s. Bonuses should be tied to quality metrics, where payment is commensurate with reaching certain thresholds of performance. For each constituent suicide prevention best practice, we recommend that quality metrics be tied to the timing and completeness of safety planning and lethal means counseling; administrative readiness and the completion of warm handoffs; and the timing, modality, and consistency of follow-up contacts. In the absence of changes to current payment and organizational structures, the Joint Commission’s NPGs will remain as recommendations. With involvement from payers and increased hospital participation in pursuing suicide prevention best practices, these recommendations are far more likely to become a reality.