Exciting new models are emerging for psychiatric consultation in the general medical hospital. In the traditional model of hospital care, consultation psychiatrists wait for a request from a medical team, perform a standard consultation, and leave their recommendations in the medical chart. This system has obvious shortcomings. For example, medical teams may not recognize which cases benefit most from consultation and may request consults for sadness and other issues that are not efficiently addressed in the hospital. Medical teams may focus initially on medical issues and only request a consultation later in the admission. Medical teams may not notice and respond to recommendations in a timely manner, and sometimes, the consultation situation simply calls for an outpatient referral, whereupon the entire process of assessment is repeated independently. This often leads to medical teams becoming frustrated with their psychiatric consultants, and psychiatric consultants becoming frustrated by inappropriate consultations.
In this article, we discuss how hospitals are experimenting with new models of consultation. Three themes characterize this expanding field. First, new models of consultation take a proactive approach in which a screening process is used to identify appropriate consultations and to identify them early in the admission. Second, new approaches emphasize embedded consultations within the medical setting, facilitating close and continuing interaction with the medical team. Third, new models are multidisciplinary, coordinating the work of social workers, advanced practice nurses, and other staff in providing psychosocial care. We argue that psychiatric consultation in the general hospital will be much different in the future.
Classic Studies Show the Promise of Hospital Consultation
We recently reviewed the long history of outcomes research in hospital consultation.1 In this area of research, length of stay (LOS) has been the most studied outcome variable. Timely and effective care produces shorter admissions, and LOS has been a partial proxy for quality of patient care. This history begins in 1934 when E. G. Billings founded the Psychiatric Liaison Department in the Colorado General Hospital, which was funded by a grant from the Rockefeller Foundation.2 During that period, Billings decreased the LOS of inpatients with psychiatric problems by 13.1 days compared to the first year of his service, when he presumably did not provide service efficiently.2 He was proud to note a savings of approximately $43 per patient. Billings noted that an important component of the improvement was a shift toward earlier consultation by 5.7 days. Several later studies reinforced the potential of psychiatric consultation to improve patient care. In 1981, Levitan and Kornfeld3 reported the effect of psychiatric consultation in female geriatric patients with femur fracture. A consultation psychiatrist followed all patients during a 6-month intervention and demonstrated an improvement in median LOS by 12 days compared to similar patients in a similar period in the previous year. In 1991, Strain et al.4 published a study of geriatric patients in two hospitals admitted with femur fracture. In the intervention year, every patient was seen within 3 days of admission by a consultation psychiatrist, who continued to participate in meetings with the surgical teams, families, and discharge planning staff during the admission. In both hospitals, LOS was reduced by 2.2 and 1.7 days, respectively.
An important study was offered by Levenson et al.5 in 1992 in a large-scale trial funded by the National Institutes of Health. All admitted patients were screened with an instrument measuring anxiety, depression, confusion, and pain to identify patients meriting consultation. During a 6-month baseline period and a 15-month intervention period, 48% of admitted patients were found to be eligible. During the intervention period, all participants screening in with this instrument were randomized to automatic consultation (n = 256) or to consultation as usual (n = 253). The authors were unable to demonstrate a statistically significant decrease in LOS. The authors did, however, perform a detailed analysis and offered a number of important insights as to why the study failed. The screening method may not have identified patients whose LOS would benefit from a psychiatric consultation. For example, patients with issues such as anxiety or depression might be better served by an outpatient referral than an inpatient consultation. The consultations performed were general in nature and not necessarily focused on the interventions that would most affect LOS. The consultations were unrequested and the authors noted the medical teams did not necessarily follow the consultant's plan. Unfortunately, the intervention group happened to be more ill, and statistical techniques are limited in their ability to analyze a trial that is unbalanced at randomization. The trial was conducted hospital-wide with a large range of LOS, further hampering the likelihood of finding a statistically significant effect. So although a meticulously designed clinical trial may fail to demonstrate its endpoint hypothesis, it can generate critical insights for further research.
In our review,1 we discuss many other studies of hospital psychiatric consultation and argue that the bulk of published studies support that psychiatric consultation is effective to the extent to which it identifies appropriate patients and provides close interaction with medical teams. We also note that research in hospital geriatric consultation in the 1980s and 1990s converged on similar conclusions. A number of studies analyzed the effects of geriatric consultation, including eight large, well-designed, randomized controlled trials in older patients, and our analysis1 concluded that successful effects on LOS were related to correctly identifying the target population, focusing on hospital issues affecting LOS, and on close collaborative relationships with the medical team. Simply doing more consultations is not enough.
New Models of Consultation Are Proactive, Embedded, and Multidisciplinary
With this background in mind, we studied the effect of a proactive approach to psychiatric consultation on a general medical unit at Yale New Haven Hospital.6 A 5-week intervention period was compared to five similar periods before and after. During the intervention period, a psychiatrist met each weekday with the medical team to review cases. In pilot work, we found that the medical team was most familiar with the patient on the day after admission, being not fully familiar with the patient on the day of admission. If necessary, the psychiatrist could examine the patient briefly or review the medical record. Based on this assessment, approximately half of the total admissions were judged to have a psychiatric issue: approximately 20% of the total admissions were judged to have psychiatric issues requiring consultation to avoid possible delay of discharge, 5% judged to have psychiatric issues requiring consultation but not likely to affect LOS, 10% were judged to have psychiatric issues that could be managed by the medical team (possibly with coaching from a consultant psychiatrist), and 15% were judged to have psychiatric issues best managed in the outpatient setting.6 Screening was quick, averaging approximately 3 minutes per admission. During the intervention period, 22% of admissions received a psychiatric consultation compared to 11% during control periods (already a relatively high rate). Consultations were performed earlier than in the control periods (1.4 vs 3 days after admission). During the intervention period, LOS improved by approximately 0.9 days. Alternatively, the number of admissions with LOS longer than 4 days was reduced (14.5% vs 27.9%). Feedback from the medical team at the end of the project complemented the ease of contacting and collaborating with the consultant, compared to the traditional paradigm in which the medical team must contact the appropriate member of a consultation team covering the entire hospital. It was apparent that substantial discussion and advice occurred regarding patients who were not actually seen by the consultant psychiatrist, although the study was not designed to quantitate such curbside consultation.6
Based on the results of this study, the Yale New Haven hospital funded a large study using a multidisciplinary proactive approach.7 An 11-month intervention period (n = 5,391 patients) was compared to a similar period in the preceding year (n = 5,158 patients). During the intervention period, a psychiatrist, psychiatric liaison nurse, and social worker provided proactive consultation to three internal medicine units. In this model, the consultation team did not regularly attend daily medical rounds, but screened cases by brief review of the medical record. The team was present on a daily basis on the units and able to receive reports easily from medical, nursing, social work, and discharge coordinators. In this model, the consultation team was able to provide a range of interventions, which could be standard psychiatric consultation, curbside consultation, direct advice to nursing staff, collaborative social work, or input to discharge planning. LOS in patients requiring psychiatric services was significantly reduced by 0.64 days, and LOS in all patients reduced by 0.3 days (analysis limited to patients with LOS <31 days). A detailed financial analysis indicated a return on investment of 1.7 to 1.8 These results prompted the hospital to fund a multidisciplinary proactive system to cover all of the internal medicine units. This team included psychiatrists, advanced practice nurses, social workers, and psychiatric liaison nurses in a structured system. This behavioral intervention team model has been successfully operating for more than 2 years. Again, medical and nursing staff have noted the improved interaction with the behavioral health team, and consultant psychiatrists and other mental health care providers have appreciated being able to identify the most appropriate cases for intervention.
Muskin et al.9 also obtained striking results with a different model of proactive psychiatric care at the Columbia Presbyterian Hospital. In their model, a psychiatrist accompanied the medical team on their daily rounds. The psychiatrist directly managed cases and wrote orders, so the approach can be described as co-management. The project involved a patient population managed by a hospitalist medical service, which is a population expected to have a briefer LOS. In this model, the consultant psychiatrist could identify cases requiring psychiatric consultation and immediately initiate it. The mean time from admission to consultation decreased by approximately 4 days. There was a mean reduction in LOS of 1.2 days compared to the previous year. Alternatively, the percentage of cases with LOS longer than 4 days decreased from 59% to 34%.9 The analysis was limited to patients with LOS less than 11 days (there was limited effect on patients with long LOS, as in our experience). An advantage of this model was the potential for a powerful role in educating the medical staff and trainees about psychiatric issues in the general hospital. In response to these results, the hospital provided funding for an additional psychiatrist and a full-time social worker to work on additional medical units. Financial modeling indicated savings due to this psychiatric intervention was more than three times the cost of salary and benefits for staff involved (ie, a return on investment more than 2-fold, similar to our results above).
Other hospitals are also experimenting with proactive models of psychiatric consultation. A multidisciplinary proactive model based on the behavioral intervention team approach was tested at the Dartmouth Hitchcock Medical Center. A 9-month pilot trial on two hospitalist medical units demonstrated significant improvements in LOS (Christine Finn, MD, written communication, February 2017). The results led to an expansion of the program to all adult units of the hospital. The current model employs effort from psychiatrists, advanced practice nurses, social workers, and psychiatric liaison nurses. As in the work described at Yale and Columbia, the team was able to provide a range of interventions besides a standard psychiatric consultation. A novel screening method uses an electronic algorithm that examines history, medications, diagnoses, order sets, and nursing care plans in the electronic medical record for all newly admitted patients to identify patients with potential psychiatric needs.
A trial is underway on three general medical units at Johns Hopkins Hospital using a psychiatrist, nurse practitioner, and social worker, with an organization similar to the behavioral intervention team model (Pat Triplett, MD, written communication, February 2017). Improved satisfaction with psychiatric services on the part of medical providers and nursing staff, as well as a substantial decrease in constant-observation staff expense, have been demonstrated in comparison to a pre-intervention period. No data on LOS or related outcome metrics are available as yet.
At Stony Brook University Medical Center, an embedded psychiatrist worked with one stable medical unit for 1 year (Brian Bronson, MD, written communication, February 2017). Clear improvements in LOS, compared to consultation as usual in the previous year, convinced the hospital to expand the model to additional medical units.
A multisite randomized trial of proactive versus usual psychiatric consultation for medical inpatients age 65 years or older is being conducted in Great Britain. It is funded by the UK National Institute of Health Research and led by Michael Sharpe, MD, of Oxford University (Michael Sharpe, MD, written communication, February 2017). Various models are being tried in different centers. There is no consensus yet on the most effective system, but Table 1 presents some suggestions for important features in any proactive approach. A systematic approach to designing collaborative systems in medical settings is available.10
Ten Hints for Effective Proactive Hospital Consultation
Recent research in proactive geriatric medical consultation has also demonstrated important effects on LOS. A study of proactive geriatric consultation by Sennour et al.11 used a geriatrician and a nurse practitioner who saw all patients age 69 years and older with physical or mental impairment. The team attended daily rounds and generally saw patients on the first day of admission. A multidisciplinary and collaborative approach is described in detail in their publication. Early involvement in discharge planning and continuing follow up on recommended interventions were clear priorities. This model of proactive geriatric consultation resembles the successful models of proactive psychiatric consultation described above. The study found a decrease of LOS of approximately 1 day, as well as improvements in other financial indicators, in a 1-year period in comparison to similar patients not enrolled in the intervention. Cost savings easily outweighed the cost of the consultation personnel. Use of predicted LOS and case mix index from the Centers for Medicare and Medicaid Services were used in an elegant analysis.11 A study by Harari et al.12 describes a differently structured but no less interesting model of systematic proactive geriatric consultation, the Older Person's Assessment and Liaison team approach. All patients age 70 years and older received a comprehensive assessment in the first 24 hours of admission compared to consultation as usual, which generally occurred several days into the admission. During the intervention, patients could receive geriatric consultation, facilitated discharge planning, and, if necessary, transfer to a geriatric unit. A significant decrease of LOS by approximately 4.1 days was observed. Other successful approaches to geriatric consultation and collaboration in orthopedic settings are described in our review.1 A study of proactive consultation for geriatric trauma patients described a 4-day improvement in LOS. This improvement was not statistically significant,13 but this may be a patient population with a high rate of behavioral issues, meriting further work.
Latest Research Expands the Model of Hospital Consultation
Although these studies illustrate the possibilities of proactive consultation, new studies are exploring expanded models of care integration. One area of interest has been coordination with outpatient care. For example, Huffman et al.14 performed a randomized prospective trial of a collaborative management of depression for patients hospitalized with acute cardiac disease. Patients screening positive for depressive symptoms on a structured assessment were randomized to usual care or the intervention. Patients in the intervention group received a systematic consultation regarding depression, including a treatment plan communicated to the inpatient and outpatient providers. During the 12 weeks after discharge, patients received three assessments of depressive symptoms. Intervention patients with continued symptoms received a reassessed care plan that was communicated to the primary care physician, and a care manager attempted to facilitate appropriate care. Patients in the intervention group had significantly improved scores in depression and other mental health outcomes compared to patients in the usual-care group. This research group extended these observations in a second randomized, controlled trial for patients hospitalized with acute cardiac conditions.15 This trial screened all patients for major depression, panic disorder, and generalized anxiety disorder, and featured an efficient intervention by telephone follow-up. Cognitive-behavioral therapy was provided as needed in the hospital and by phone. The results indicated statistically significant superiority in quality of life and other outcome scores compared to usual care.
A project with a different approach to coordinating care after discharge for patients with sickle cell disease (SCD) was carried out at our hospital (Ariadna Forray, MD, written communication, January 2017). The hospital made a major commitment to development of a coordinated inpatient and outpatient SCD program. The multidisciplinary program consists of a clinical director, two advanced practice nurses, two social workers, a psychiatrist, and a dedicated inpatient unit and support services. Key components of the program included structured inpatient pain-management strategies; continuity of care between inpatient and outpatient services; screening for depression, anxiety, and substance use; co-localized mental health treatment; and community outreach. A single psychiatrist saw all SCD patients in the hospital for consultation, and the same psychiatrist also provided collaborative care in the SCD outpatient clinic. This systematic longitudinal integration of care, together with assignment of specialist hematological and psychiatric care providers, resulted in a dramatic improvement in LOS and costs, as well as other patient care metrics (Ariadna Forray, MD, personal communication). In our review of consultation interventions,1 we noted that units or pathways specialized for specific diagnostic categories produced the greatest effect in LOS and other outcome metrics.
Another novel approach to integrating hospital consultation with other services was offered by Zaubler et al.16 The Hospital Elder Life Program is multicomponent delirium-prevention program that includes assessment by an elder life specialist, daily visits, reduction of potentially deliriogenic medications, assistance with feeding and drinking, cognitive stimulation, maintenance of sleep periods, and other interventions previously shown to decrease the incidence of delirium in geriatric patients.17 These investigators adapted the program for a community hospital. A distinctive feature of their approach was the use of volunteers recruited from the community. Delirium incidence was significantly reduced by 8%, and LOS was significantly reduced by 2.2 days during a 9-month intervention period compared to a 4-month control period prior to the intervention.17
Psychiatric Consultation in the General Hospital Will Be Different in the Future
Substantial research has shown the power of integrated, collaborative behavioral health care in outpatient medical clinics.18 In these models, all patients are screened systematically for behavioral health issues. Although different models depend on differing methods of interaction, all models implement a structured and close coordination of medical and psychiatric care. These new models generally include multidisciplinary collaboration, which uses the skill sets of different providers in a cost-effective way. Such integrated models have been shown to produce improvements in multiple outcome measures. We would highlight the increased satisfaction for both consultant and consulting teams with improved ease of communication and improved focus on appropriate cases. In this brief review, we have attempted to highlight similar trends in new approaches to hospital consultation that display the effectiveness of proactive, embedded, and multidisciplinary approaches. Closer integration with outpatient services and more structured approaches to specific disease categories will be important new frontiers. Incentives for avoiding readmission to the hospital will increase the importance of coordination with the outpatient domain. Better coordination of care for complex patients across outpatient, emergency department, and medical hospital is inevitable. We predict that psychiatric consultation in the medical hospital will change radically in the next decades. When E. G. Billings began the Psychiatric Liaison Department of the Colorado General Hospital in 1934, he meant to institute a system that would teach the hospital to recognize and address the psychiatric issues of all patients.19 His service offered teaching, coordination, and supervision beyond simple consultation. His service included a social worker from the start, and was thus multidisciplinary. The new collaborative models we have discussed emphasize the “liaison” aspect of consultation-liaison psychiatry, and represent nothing more than a return to the roots of the field.
- Desan PH, Zimbrean PC, Lee HB, Sledge WH. Proactive consultation services for the general hospital of the future. In: Summergrad P, Kathol RG, eds. Integrated Care in Psychiatry. New York, NY: Springer Verlag; 2014:157–181. doi:10.1007/978-1-4939-0688-8_10 [CrossRef]
- Billings EG. Value of psychiatry to the general hospital. Hospitals. 1941;15:30–34.
- Levitan SJ, Kornfeld DS. Clinical and cost benefits of liaison psychiatry. Am J Psychiatry. 1981;138(6):790–793. doi:10.1176/ajp.138.6.790 [CrossRef]
- Strain JJ, Lyons JS, Hammer JS, et al. Cost offset from a psychiatric consultation-liaison intervention with elderly hip fracture patients. Am J Psychiatry. 1991;148(8):1044–1049. doi:10.1176/ajp.148.8.1044 [CrossRef]
- Levenson JL, Hamer RM, Rossiter LF. A randomized controlled study of psychiatric consultation guided by screening in general medical inpatients. Am J Psychiatry. 1992;149(5):631–637. doi:10.1176/ajp.149.5.631 [CrossRef]
- Desan PH, Zimbrean PC, Weinstein AJ, Bozzo JE, Sledge WH. Proactive psychiatric consultation services reduce length of stay for admissions to an inpatient medical team. Psychosomatics. 2011;52(6):513–520. doi:. doi:10.1016/j.psym.2011.06.002 [CrossRef]
- Sledge WH, Gueorguieva R, Desan PH, Bozzo JE, Dorset J, Lee HB. Multidisciplinary proactive psychiatric consultation service: impact on length of stay for medical inpatients. Psychother Psychosom. 2015;84(4):208–216. doi:. doi:10.1159/000379757 [CrossRef]
- Sledge WH, Bozzo J, White-McCullum BA, Lee HB. The cost-benefit from the perspective of the hospital of a proactive psychiatric consultation service on inpatient general medicine services. Health Econ Outcome Res. 2016;2(4):122. https://www.omicsonline.org/open-access/the-costbenefit-from-the-perspective-of-the-hospital-of-a-proactivepsychiatric-consultation-service-on-inpatient-general-medicine-.php?aid=81448. Accessed May 24, 2017.
- Muskin PR, Skomorowsky A, Shah RN. Co-managed care for medical inpatients, C-L vs C/L psychiatry. Psychosomatics. 2016;57(3): 258–263. doi:. doi:10.1016/j.psym.2016.02.001 [CrossRef]
- Beach SR, Walker J, Celano CM, Mastromauro CA, Sharpe M, Huffman JC. Implementing collaborative care programs for psychiatric disorders in medical settings: a practical guide. Gen Hosp Psychiatry. 2015;37(6): 522–527. doi:. doi:10.1016/j.genhosppsych.2015.06.015 [CrossRef]
- Sennour Y, Counsell SR, Jones J, Weiner M. Development and implementation of a proactive geriatrics consultation model in collaboration with hospitalists. J Am Geriatr Soc. 2009;57(11):2139–2145. doi:. doi:10.1111/j.1532-5415.2009.02496.x [CrossRef]
- Harari D, Martin FC, Buttery A, O'Neill S, Hopper A. The older persons' assessment and liaison team ‘OPAL’: evaluation of comprehensive geriatric assessment in acute medical inpatients. Age Ageing. 2007;36(6):670–675. doi:. doi:10.1093/ageing/afm089 [CrossRef]
- Lenartowicz M, Parkovnick M, McFarlan A, et al. An evaluation of a proactive geriatric trauma consultation service. Ann Surg. 2012;256(6):1098–1101. doi:. doi:10.1097/SLA.0b013e318270f27a [CrossRef]
- Huffman JC, Mastromauro CA, Sowden G, Fricchione GL, Healy BC, Januzzi JL. Impact of a depression care management program for hospitalized cardiac patients. Circ Cardiovasc Qual Outcomes. 2011;4(2):198–205. doi:. doi:10.1161/CIRCOUTCOMES.110.959379 [CrossRef]
- Huffman JC, Mastromauro CA, Beach SR, et al. Collaborative care for depression and anxiety disorders in patients with recent cardiac events: the Management of Sadness and Anxiety in Cardiology (MOSAIC) randomized clinical trial. JAMA Intern Med. 2014;174(6):927–935. doi:. doi:10.1001/jamainternmed.2014.739 [CrossRef]
- Zaubler TS, Murphy K, Rizzuto L, et al. Quality improvement and cost savings with multicomponent delirium interventions: replication of the Hospital Elder Life Program in a community hospital. Psychosomatics. 2013;54(3):219–226. doi:. doi:10.1016/j.psym.2013.01.010 [CrossRef]
- Inouye S, Inouye D, Baker P, Fugal E, Bradley EH. Dissemination of the Hospital Elder Life Program: implementation, adaptation, and successes. J Am Geriatr Soc. 2006;54(10):1492–1499. doi:. doi:10.1111/j.1532-5415.2006.00869.x [CrossRef]
- Archer J, Bower P, Gilbody S, et al. Collaborative care for depression and anxiety problems. Cochrane Database Syst Rev. 2012;10: CD006525. doi:10.1002/14651858.CD006525.pub2 [CrossRef].
- Billings EG. The psychiatric liaison department of the University of Colorado medical school and hospitals. Am J Psychiatry. 1966;122(12 Suppl):28–33. doi:10.1176/ajp.122.12S.28 [CrossRef]
Ten Hints for Effective Proactive Hospital Consultation
Targeting the right population. The consultation system must identify the cases that will benefit specifically from in-hospital services, rather than cases with any psychiatric issue.
Efficient screening. The consultation system must have a rapid and efficient method for screening patients, but an in-person evaluation of all admitted patients is costly. In published research, close interaction with the medical, nursing, social work, and care coordination teams has been effective, as has targeted review of the medical record, especially medical team sign outs.
Electronic medical records. An electronic record is essential to efficient screening, communication with the team, and monitoring of patient progress; truly integrated consultation is much harder with a paper chart.
Embedding of psychiatric services. Dedicating specific personnel to specific hospital units and inserting them in specific rounds increases effectiveness. Communication occurs efficiently when providers are in regular contact. Actual co-management is the fullest form of embedded services.
Informal consultation. Part of the impact of proactive and embedded consultation seems to be informal curbside consultation, although no research has yet measured such interaction quantitatively. Such consultation might be especially critical in targeting discharge planning from early in admission.
Flexibility of response. The efficient integrated consultation team of the future will offer a range of responses, including informal curbside consultation, consultation with an advanced practice registered nurse or physician assistant provider, social work interventions, advice to discharge planning teams, psychological services, and specialized providers such as addiction teams, as well as traditional consultation with an experienced psychiatrist.
Implementation of consult recommendations. Multiple authors have noted that the relationship with medical teams is critical and may require development over time as the integrated consultant will have to prove his or her usefulness. Medical teams are more likely to implement interventions when the consultant is in regular contact.
Close follow-up. In new paradigms, the consultant role is an ongoing one, not a single consultation interaction. The consultant must be able to monitor success of interventions, and be able to detect and react to new needs.
Outpatient liaison. Many psychiatric issues encountered in the hospital are appropriately handled in the outpatient setting. The consultation team must have a process for efficient referral to the correct community resources. This referral should include effective channels for information sharing.
Integration of care. The model we are describing here extends beyond consultation. The consultation service of the future will be part of truly integrated care, with health care providers of different disciplines with specific protocols and pathways for different types of patients, with specific units and teams for diverse subgroups of patients such as patients with addictions, and with coordination of care across different settings.