The length of inpatient stay has substantially decreased in the U.S.

Psychiatric inpatient treatment has declined drastically in the U.S. over the past several decades, starting with the deinstitutionalization movement of the 1960s and 1970s. This movement was motivated by a combination of factors, including concern over inhumane conditions in long-term care facilities, a shifting philosophy towards less-restrictive mental health treatment, and the emergence of effective psychotropic medications that enabled care for many individuals with severe mental illness to be managed in community settings (e.g., outpatient clinics, partial hospitalization programs). Several key policy changes over the past 50 years have redirected federal funding from hospitals to community-based care, starting with President Kennedy’s 1963 Community Mental Health Act and the enactment of Medicaid in 1965 (Lutterman, Reference Lutterman2022; Salinsky & Loftis, Reference Salinsky and Loftis2007). Similar deinstitutionalize movements were happening internationally (Fakhoury & Priebe, Reference Fakhoury and Priebe2002) with the enactment of similar policies, like the Mental Health Act of 1959 in the U.K., which removed the distinction between psychiatric and other hospitals (Turner, Reference Turner2004), the 1978 Basglia Law in Italy mandating the closure of all psychiatric hospitals (Badano, Reference Badano2024), the Psychiatric Reform Law of 2001 in Brazil that expanded community-based services (Taborda, Reference Taborda2013), and the Mental Health Care Act of 2002 in South Africa emphasizing treatment in the least restrictive environment (Szabo & Kaliski, Reference Szabo and Kaliski2017). Since 1970, the number of individuals in inpatient and residential care beds on a given day in the U.S. has reduced by over 60%, from 471,451 to 187,877 in 2018, which was largely driven by decreases in state and county psychiatric hospital beds (−90.3%), while private psychiatric hospital capacity increased (+396.2%). Despite increases in some settings, accounting for population growth, the overall rate of patients in inpatient care declined by 76% from 236.8 to 56.8 per 100,000 between 1970 and 2018 (Lutterman, Reference Lutterman2022).

Reductions in capacity and increased emphasis on less restrictive community-based treatment led to a substantial decrease in the duration of hospitalization (Lave, Reference Lave2003). The average length of hospital stay shortened drastically from long-term care lasting years/decades prior to the 1950s to hospital stays lasting weeks/months in the latter half of the 20th century. The rise in managed care starting in the 1980s applied further pressure to shorten the length of hospital stays to reduce costs and increase patient turnover (Geller, Fisher, McDermeit, & Brown, Reference Geller, Fisher, McDermeit and Brown1998; Lave, Reference Lave2003). Although variable depending on the setting, region, and patient characteristics, estimates for the current average length of inpatient stays range from 3 to 7 days (Adepoju, Kim, & Starks, Reference Adepoju, Kim and Starks2022; Shah, Leontieva, & Megna, Reference Shah, Leontieva and Megna2020), which is a reduction even relative to the early 2000s (~10 days; Lave, Reference Lave2003; Tulloch, Fearon, & David, Reference Tulloch, Fearon and David2011). The current inpatient care model in the U.S. is a new age of ultra-brief hospital stays focused on crisis stabilization and rapid movement to lower levels of care (Glick, Sharfstein, & Schwartz, Reference Glick, Sharfstein and Schwartz2011; Sharfstein, Reference Sharfstein2009). This shift has forced providers to adapt psychological treatments to fit within a timeline they are not designed for without relevant research to inform these changes (Snyder, Clark, & Jones, Reference Snyder, Clark and Jones2012).

Group therapy is the predominant mode of psychological treatment in inpatient care

Inpatient hospitalization is the most expensive form of mental health treatment, and the majority of these costs are attributable to staffing (Salinsky & Loftis, Reference Salinsky and Loftis2007). Different from other acute medical care settings where expensive equipment, materials, or space (e.g., imaging machines, laboratory testing, operating room suites) may make up a substantial portion of costs, staff time accounts for over 85% of inpatient psychiatric treatment costs (Cromwell et al., Reference Cromwell, Maier, Gage, Drozd, Osber, Richter and Goldman2004). This time involves activities such as assessment, medication management, psychotherapy, and behavioral monitoring to ensure safety. In outpatient treatment, these activities are paid on a per-service basis (e.g., per therapy hour). However, as a result of the Medicare, Medicaid, and SCHIP (Supplemental Children’s Health Insurance Program) Balanced Refinement Act of 1999’s mandate to implement a per-diem reimbursement rate for inpatient psychiatric stays (Balanced Budget Refinement Act, 1999), inpatient treatment is now generally covered at a per-diem rate across providers, which incentivizes prioritization of services that are efficient with respect to staff time. Group therapy is a particularly efficient service because multiple patients can receive treatment from a single provider simultaneously.

The per-diem payment system, coupled with policy mandates pressuring state hospitals to increase treatment offered during inpatient stays, has led to the current norm in which group therapy is the main modality of psychological treatment in inpatient units, and access to individual therapy is limited (Cook, Arechiga, Dobson, & Boyd, Reference Cook, Arechiga, Dobson and Boyd2014; Scaturo, Reference Scaturo2004). For example, in 2003, state hospitals in North Carolina were ordered to increase the minimum hours of active treatments provided to all patients to 20 hours/week and responded by ramping up group programming so that patients were spending several hours/day in therapy groups (Snyder, Clark, & Jones, Reference Snyder, Clark and Jones2012). Although there have been recent efforts to develop ultra-brief psychological EBTs for inpatient care, these treatments have almost exclusively been individual therapy models (e.g., Diefenbach et al., Reference Diefenbach, Lord, Stubbing, Rudd, Levy, Worden and Everhardt2024; Paterson et al., Reference Paterson, Karatzias, Dickson, Harper, Dougall and Hutton2018; Tyrberg & Klintwall, Reference Tyrberg and Klintwall2022). As we detail below, research on inpatient psychological treatment in general is very limited, but for inpatient groups, this is even more so the case.

Inpatient staff delivering psychological treatment have limited training in EBT

Another consequence of the cost-conscious managed care era of U.S. healthcare is that hospitals are incentivized to hire staff with the lowest credentials (and, thus, who can be paid the lowest rate) who can provide “equivalent” services. Lack of research leads to ambiguity about the level of training necessary for treatment to be implemented in a way that it can reach a standard of evidence-based care. Consequently, inpatient units employ few psychologists, primarily relying on masters-level clinicians (and in some cases, bachelor-level clinicians) to provide psychological services, who are likely not trained to deliver the type of EBTs that currently exist (Berry et al., Reference Berry, Raphael, Haddock, Bucci, Price, Lovell, Drake, Clayton, Penn and Edge2022; Dandan, Mansour, & Diab, Reference Dandan, Mansour and Diab2024; Pudalov, Swogger, & Wittink, Reference Pudalov, Swogger and Wittink2018). A recent systematic review of staff perspectives on the barriers and facilitators to the implementation of psychological treatments in inpatient settings identified a lack of staff training in psychological EBTs as one of the main barriers (Evlat, Wood, & Glover, Reference Evlat, Wood and Glover2021). This is especially the case for group therapy. Irvin Yalom, who is widely considered an authority on group therapy and has written one of the only clinical texts on inpatient group therapy, cautions that lack of hospital leadership buy-in to the value of group therapy leads to a number of consequences that undermine the effectiveness of groups (e.g., inconsistent scheduling, frequently pulling members out of the group to complete administrative tasks), including assigning groups to be led by staff with the least experience with psychological treatments (e.g., nurses; Yalom, Reference Yalom1983; Yalom & Leszcz, Reference Yalom and Leszcz2020).

Across settings, clinical trials of psychological treatments overwhelmingly rely on psychologists and doctoral trainees to administer treatments. Research on inpatient group therapy is no exception, and the vast majority of inpatient therapy research has involved group interventions implemented by Ph.D.-level psychologists (e.g., Boynton & Sanderson, Reference Boynton and Sanderson2022; Esposito-Smythers, McClung, & Fairlie, Reference Esposito-Smythers, McClung and Fairlie2006; Heriot-Maitland, Vidal, Ball, & Irons, Reference Heriot-Maitland, Vidal, Ball and Irons2014; Raune & Daddi, Reference Raune and Daddi2011). The norm of masters- or bachelors-level clinicians and other staff (e.g., nurses) providing group therapy on inpatient units is unlikely to change without major shifts in financial incentives or regulations on which disciplines can provide group therapy. We are not recommending such regulations be implemented; rather, we argue that clinical trials of inpatient group therapy should prioritize utilizing staff who are typically responsible for running groups in the real world. Given the often lack of background in psychological EBTs of non-psychologist staff, training may be a necessary prerequisite. Research on training in psychological EBTs is itself very limited (Frank, Becker-Haimes, & Kendall, Reference Frank, Becker-Haimes and Kendall2020). Training multidisciplinary staff to deliver psychological EBTs is another area of research in dire need of development and is virtually nonexistent for the implementation of inpatient group therapy.

Research on ultra-brief inpatient group therapy is extremely limited

Research on inpatient group therapy had a period of proliferation in the 1970s/80s, following a trend of increased interest in group therapy more broadly. This research has been summarized in meta-analyses (Burlingame, Fuhriman, & Mosier, Reference Burlingame, Fuhriman and Mosier2003; Kösters, Burlingame, Nachtigall, & Strauss, Reference Kösters, Burlingame, Nachtigall and Strauss2006), which supported the effectiveness of inpatient group therapy, though effect sizes were notably smaller compared to outpatient groups (Burlingame, Fuhriman, & Mosier, Reference Burlingame, Fuhriman and Mosier2003). Even though they were published 20 years ago, both meta-analyses noted an already significant decline in inpatient group research in the U.S. since the 1980s, while research increased in other countries. For example, 60% of the studies included in Kösters, Burlingame, Nachtigall, and Strauss’s (Reference Kösters, Burlingame, Nachtigall and Strauss2006) meta-analysis were published in Germany, where the average length of stay is >5 weeks (Dimitri et al., Reference Dimitri, Giacco, Bauer, Bird, Greenberg, Lasalvia, Lorant, Moskalewicz, Nicaise, Pfennig, Ruggeri, Welbel and Priebe2018). This trend has only continued in recent years, with the vast majority of contemporary inpatient group research conducted outside the U.S., mostly in Germany (Weber & Strauss, Reference Weber and Strauss2015). Much of this research is on psychodynamic-oriented groups (Radcliffe & Diamond, Reference Radcliffe and Diamond2007; Strauss & Schmidt, Reference Strauss and Schmidt2020), which is the dominant treatment model in Germany and was the primary model of inpatient group therapy in the U.S. in the 1970s/80s, whereas cognitive-behavioral therapy (CBT) has become the main treatment model in the U.S. in recent decades (Emond & Rasmussen, Reference Emond and Rasmussen2012). Consequently, past research in the U.S. and more recent research conducted in other countries has limited relevance to contemporary inpatient treatment in the U.S., where the average length of stay is magnitudes shorter than the group interventions utilized in these studies.

Length of stay is not the only factor rendering much of the research on inpatient group therapy irrelevant to contemporary inpatient treatment in the U.S. The vast majority of studies involved therapy groups that were homogeneous with respect to diagnosis and/or severity, with several studies conducted on specialized units (e.g., for substance use or eating disorders), and followed a closed-group structure (i.e., all members enter group on the same day and complete the series of group sessions together; Cook, Arechiga, Dobson, & Boyd, Reference Cook, Arechiga, Dobson and Boyd2014; Emond & Rasmussen, Reference Emond and Rasmussen2012). In practice, inpatient groups tend to be highly heterogeneous both with respect to diagnosis and level of functioning. In the standard 24-bed psychiatric unit, there are simply not enough patients to split up into multiple specialized groups (Yalom, Reference Yalom1983). Additionally, ultra-short stays and quick patient turnover make it nearly impossible to have closed multi-session therapy groups with the same members (Emond & Rasmussen, Reference Emond and Rasmussen2012). Differences in the timing of admission and duration of stay mean that, on any given day, a different composition of patients will attend group, and individual patients will receive different exposures to group (different duration and/or modules). This is so much the case that Yalom suggests that inpatient group leaders “assume that your group will last for only a single session” (Yalom & Leszcz, Reference Yalom and Leszcz2020, p. 596), which is echoed in others’ calls for a stand-alone session format for inpatient groups (Cook, Arechiga, Dobson, & Boyd, Reference Cook, Arechiga, Dobson and Boyd2014; Heriot-Maitland, Vidal, Ball, & Irons, Reference Heriot-Maitland, Vidal, Ball and Irons2014; Raune & Daddi, Reference Raune and Daddi2011; Snyder, Clark, & Jones, Reference Snyder, Clark and Jones2012). A survey of inpatient psychologists found that group heterogeneity and rapid turnover were among the main barriers to adapting evidence-based treatments for the inpatient setting (Snyder, Clark, & Jones, Reference Snyder, Clark and Jones2012).

The very small literature from the past 20 years that does exist serves to engender optimism for the benefits of group therapy adapted from outpatient EBTs for ultra-brief inpatient care. First, a few studies have found positive results for patient satisfaction with inpatient groups based on CBT, specifically in heterogeneous, rolling attendance groups (Boynton & Sanderson, Reference Boynton and Sanderson2022; Esposito-Smythers, McClung, & Fairlie, Reference Esposito-Smythers, McClung and Fairlie2006; Heriot-Maitland, Vidal, Ball, & Irons, Reference Heriot-Maitland, Vidal, Ball and Irons2014; Nikolitch et al., Reference Nikolitch, Laliberté, Yu, Strychowsky, Segal, Looper and Rej2016; Raune & Daddi, Reference Raune and Daddi2011). CBT is largely considered the most EBT model for briefer inpatient groups (David, Cristea, & Hofmann, Reference David, Cristea and Hofmann2018). A smaller body of research suggests that psychodynamic-oriented inpatient groups have also demonstrated positive effects in longer group formats (Radcliffe & Diamond, Reference Radcliffe and Diamond2007; Strauss & Schmidt, Reference Strauss and Schmidt2020). A few studies have examined outcomes beyond feasibility and acceptability, observing reductions in symptoms (e.g., eating pathology, distress) and hospital re-admission (Gaudiano et al., Reference Gaudiano, Ellenberg, Ostrove, Johnson, Mueser, Furman and Miller2020; Gaudiano et al., Reference Gaudiano, Ellenberg, Johnson, Mueser and Miller2023; Veltro et al., Reference Veltro, Falloon, Vendittelli, Oricchio, Scinto, Gigantesco and Morosini2006; Veltro et al., Reference Veltro, Vendittelli, Oricchio, Addona, Avino, Figliolia and Morosini2008; Wiseman, Reference Wiseman2002). However, just three of these studies involved heterogeneous, rolling attendance groups, one of which studied an intervention involving both group and individual therapies (Gaudiano et al., Reference Gaudiano, Ellenberg, Johnson, Mueser and Miller2023), and two reported on the same clinical trial conducted in Italy (Veltro et al., Reference Veltro, Falloon, Vendittelli, Oricchio, Scinto, Gigantesco and Morosini2006; Veltro et al., Reference Veltro, Vendittelli, Oricchio, Addona, Avino, Figliolia and Morosini2008), where the average hospital stay is >2 weeks (Dimitri et al., Reference Dimitri, Giacco, Bauer, Bird, Greenberg, Lasalvia, Lorant, Moskalewicz, Nicaise, Pfennig, Ruggeri, Welbel and Priebe2018). Finally, there is some precedent for training inpatient staff, and specifically nondoctoral level staff, to deliver inpatient groups based on EBTs with adequate fidelity (Gaudiano et al., Reference Gaudiano, Ellenberg, Ostrove, Johnson, Mueser, Furman and Miller2020; Gaudiano et al., Reference Gaudiano, Ellenberg, Johnson, Mueser and Miller2023; Tyrberg, Carlbring, & Lundgren, Reference Tyrberg, Carlbring and Lundgren2017). This is clearly a very limited pool of research, such that it is easier to summarize what has been studied than to identify gaps that need to be filled. No study has involved an inpatient group model that addresses all of the important dimensions that distinguish contemporary inpatient care in the U.S., such as ultra-brief stays, quick patient turnover, group heterogeneity, and reliance on nonpsychologist group leaders.