Chapter 1. The Hospitalist Model in Psychiatry

In acute psychiatric settings, clear communication is essential. Although electronic health records have made the issue of legibility less of a problem, most hospital charting fails to convey the richness and complexity of a patient’s presentation and situation. Communication with patients requires face-to-face time, and team interactions are more meaningful when they are conducted in person rather than on the telephone or through the chart. Yes, it is possible to determine when a patient receives medications, how many hours he slept, or his latest vital signs through a chart. But a qualitative difference is seen when the opportunity exists to engage in person. In a team, physicians can learn from the different disciplines: nurses know when and why patients refuse medications or take as-needed medications, activity therapists can report on how patients are functioning in tasks and interacting with others, and social workers communicate with families about their concerns. All of this information is needed to ensure that the patient gets the right treatment and is safely discharged.

Although physicians may or may not be devoted full time to the hospital, the rest of the staff at a hospital generally is. Although physicians’ prescriptions determine what kinds of medications the patient may receive or what other major intervention will occur (such as electroconvulsive therapy [ECT]), nurses, activity therapists, and social workers are the ones who spend the bulk of the time with patients. They are also the ones to whom patients and families are more likely to confide thoughts and concerns. A physician who is seldom available or who is not well known to the rest of the staff may not hear about a concern either because he or she is not around or because staff might not feel comfortable communicating that concern. Nurses, social workers, patient care workers, and activity therapists often have good instincts and observational skills. Their assessment of patients’ progress (or lack of progress) is important in understanding what interventions still need to be done and when a patient is getting close to discharge. A hospitalist psychiatrist who builds relationships with the team and is more accessible will be in a position—both physically and interpersonally—to get needed information.

Physicians will learn more when they know (and are known by) their teams, and teams will be better able to take direction from familiar physicians. Teams function best with mutual trust. If the rest of the team does not really know a physician, the nurses might unintentionally convey skepticism to patients about the treatment intervention, or the social workers might find themselves in a position of allying with families against the treating physician. There might be delays in physician orders being carried out if the orders are not understood or if the orders are not appropriate for the setting (because of lack of knowledge of the physician). Hospitalist psychiatrists can build trusting relationships with their teams that involve clear communication both internally and externally and also result in clarity of the treatment plans with patients and families.

Consistency is also important, especially for trainees. Psychiatry as a field can be fraught with disagreements in management of some conditions. A patient with mood swings on an hourly basis, who also describes interpersonal chaos, chronic feelings of emptiness, and patterns of self-harm behaviors, might be diagnosed by some psychiatrists as having bipolar disorder and others as experiencing the symptoms of borderline personality disorder. Whatever the merits of either side, trainees need to learn from faculty how to manage patients with symptoms. It is disruptive to trainees—and poor patient care—to have faculty rotate through inpatient settings with wildly different approaches to problem solving and patient presentation. If trainees are assigned to the hospital for up to a year at a time, they could experience a great deal of turmoil with rotating attendings. A smaller cadre of dedicated hospitalist psychiatrists allows more consistency of approach (while permitting individual differences in style or team interaction) to help keep trainees from lurching from one intervention style to another. Furthermore, hospitalist psychiatrists can more easily engage trainees in decision-making (because they are already fully immersed in the environment) and bedside teaching and other educational efforts.

Trainees sometimes experience inconsistent messages from faculty because clinicians have different levels of comfort with acutely ill patients. A psychiatrist who works primarily with depressed patients in an ambulatory setting might not know how to treat a patient with chronic schizophrenia who at baseline is somewhat disorganized and hears voices. The trainee who is attempting to manage this patient might hear from this faculty member that the patient needs to remain in the hospital because of the psychosis, whereas a psychiatrist who works regularly with the chronically mentally ill might have a better understanding that such patients can be supported in the community even when they have symptoms. On a consultation service, a faculty member who has comfort with how to assess suicidal patients might approach a patient with chronic suicidal ideation and self-injurious behavior differently from a faculty member who seldom encounters patients talking about suicide.

In most cases, comfort level develops in parallel with expertise based on experience with many similar cases informed by familiarity with available clinical evidence and consensus guidelines of recognized authorities in the field. This level of engagement facilitates the development of pattern recognition and intuitive understanding that constitutes mastery of any endeavor. Patients benefit from care that is provided by experienced clinicians working in an environment in which they are familiar with the diagnoses, treatments, and other issues they regularly encounter. Hospital-based care is more than just high-level outpatient care. The approaches to diagnosis and treatment; expectations of patients, family, and team members; resources available; and limitations are all unique to the hospital setting and are best handled by those with the most experience and skill with this level of care.

One of the biggest challenges in health care today is the lack of resources, in terms of both space and time and reimbursement. A well-documented benefit of hospitalist medicine is that the model increases efficiency for both admissions to and discharges from the hospital. Psychiatry faces its own problems with regard to throughput at the hospital. Scheduling outpatient appointments is often difficult, and hospital beds are often full. Insurance reimbursement rates are low, and insurance companies are increasingly insisting on rapid transition from hospital to outpatient care. In that setting, a hospital system needs to be as efficient as possible to appropriately and safely move patients from one level of care to another. A physician in the emergency department who is not familiar with psychiatric patients might decide that someone who engages in chronic self-injurious behavior is attempting suicide and needs to be hospitalized, whereas a hospital-based psychiatrist could see the same patient and understand that he would benefit most from a return to his outpatient team that is working with him on dialectical behavior therapy techniques. A rounding faculty member on an inpatient unit who does not know the community resources might be hesitant to discharge a patient to an outpatient team that will follow her. If emergency physicians fill up the scarce psychiatric hospital beds, or if occasional hospital faculty are not comfortable discharging patients, it will become even harder to move patients out of emergency departments, through hospital care, and back to the community.

Not only do psychiatrists who are more comfortable and familiar with acutely ill patients practice in a more efficient manner, but also it becomes easier to deal with insurance issues. As anyone who has worked on an inpatient unit knows, insurance companies are becoming increasingly stingy with authorizations for both admission and continued stays. There is an art to learning to communicate what we are seeing with patients to insurance companies to help them understand how we are trying to help their members. Now that danger to self or others is the major criterion for hospitalization, stating that a patient is still suicidal is not sufficient to obtain authorization. Insurance companies have started to ask more pointed questions such as whether the patient has a plan or an intent to kill himself or herself. It takes practice—and coaching—to convey concerns about patients to the entities that pay for care.

Some psychiatric services such as ECT are more efficient when managed through a hospitalist model. Although many patients can receive ECT treatments when they are living at home, the ECT treatment itself requires a brief staff check-in with a patient, general anesthesia, and recovery time. In many ways, it works much like an outpatient surgery. Within a hospital, an ECT psychiatrist has access to anesthesiology and nurse anesthetists, as well as medical backup in case of a medically unstable patient. An ECT team within a hospitalist model also can easily include inpatients from the psychiatric unit, as well as patients on the medical or surgical services who might benefit from the treatment.

When the hospitalist model was first implemented within medicine services in hospitals in the 1990s, the question was raised about how a patient might respond to the discontinuity in care. The assumption was that the hospitalist model was substituting a stranger—the specialized hospital physician—for the patient’s physician from the clinic. Concerns were also raised about whether physicians who were attuned to the fiscal and regulatory realities of the hospital would be sufficiently attentive to the needs of the patient. Those issues have been mostly resolved within medical services that use a hospitalist model. But how much are they concerns for hospitalist psychiatrists?

No systematic evidence is available on discontinuity in treatment for psychiatric patients. Many fewer patients are likely in established care relationships with a psychiatrist when they have a crisis that requires psychiatric hospitalization. Patients within the community mental health system are more likely to have episodes of instability, but their treatment teams have been designed to focus on outpatient services rather than inpatient care. For patients who present because of acute substance use issues, the challenge is getting them to engage with any type of provider. Because many outpatient psychiatrists do not have stable, long-term patient populations, relatively few patients would be significantly affected by the discontinuity of a hospitalist psychiatrist.

The potential for hospitalist psychiatrists’ divided loyalty to fiscal issues and patient care is harder to assess. Part of the issue is that the purpose of psychiatric hospitalization has changed so much over the past few decades that hospital providers may be more affected by their own experience of the value of the hospital stay than by either what the patient wants or what the insurance company will reimburse. When hospital stays were still measured in months, providers could anticipate that a patient could recover from an episode of depression in the hospital. Today, there is only time to start an antidepressant, make sure a patient is no longer suicidal, and discharge to outpatient care. The reality is that insurance reimbursement for psychiatric hospitalization is limited regardless of where a psychiatrist spends most of his or her time.

Although communication, teamwork, consistency, and efficiency are major advantages within a hospitalist model, there are definitely challenges. One of the most frequently documented concerns is the problem with handoffs to the next level of care provider. All of the rich, important interpersonal interactions that are so helpful in working with patients on inpatient teams would also be helpful in transitioning patients from inpatient to outpatient care, but those interactions are much harder to arrange. In an ideal situation, the inpatient team would know exactly where a discharged patient is headed and would be able to give a verbal handoff to the receiving provider. That ideal often goes unrealized. Some patients do not have established providers before hospitalization. In that context, it is difficult to introduce the patient and his or her issues to a new provider (if he or she could even be reached). Sometimes patients will be engaged in split treatment after discharge and will see a therapist right away while a primary care physician manages medications. Primary care physicians are often hard to reach, and it is not obvious in those cases how much detail to share or what recommendations to make for ongoing care.

Not only is it highly variable in terms of who will be seeing a patient after discharge, but also the timing of the handoff becomes complicated. Discharges have to be managed efficiently, and sometimes inpatient teams do not find out who a patient will be seeing until shortly before he or she leaves the hospital. Trying to reach outpatient providers who have busy clinic schedules is often time-consuming. If a hospital psychiatrist leaves a message for an outpatient psychiatrist about a discharging patient, the receiving physician might not be able to return the call for a few days. In the intervening time, the hospitalist psychiatrist, who has a constantly changing patient population, might lose track of details of an already-discharged patient.

Written handoffs often lack clear information. Regulations on information that has to be conveyed to third-party payers and metrics for quality sometimes mean that discharge summaries are geared toward reimbursement rather than conveying clear information. Accomplishing both with a discharge summary is certainly possible. However, there is a definite art to creating a written handoff that helps an outside provider understand what the patient was like before hospitalization, what happened during the hospital stay, and what the expectation is for treatment following discharge. Of course, the goal of a training program is to teach that art. But training programs have so much to teach residents in terms of patient care that writing discharge summaries seldom gets sustained attention, and physicians vary in their ability to communicate through medical records. The ideal, which requires practice, is to have both written and verbal communication in patient handoffs. This topic is addressed in detail in Chapter 12, “Transitions in Care, Documentation, and Interdisciplinary Communication.”

Another disadvantage of a hospitalist model is that hospital-based psychiatrists can focus so much on acute issues and management in the structured setting that they can lose sight of how patients need to function in their outside lives. This can affect how hospital psychiatrists prescribe medication, for example. In the hospital, it does not matter whether a patient takes a medication three times a day or once—a nurse is available to dispense it. At home, however, it is seldom realistic to expect that patients will administer their own medications with the same thoroughness. Medication selection is also more complicated in a transition to outpatient care. Hospital formularies usually do not make distinctions between levels of cost for medications; either the hospital will allow providers to prescribe medications or they will not. But the medications that are covered and available in the hospital are not necessarily the ones covered by a patient’s outpatient pharmacy benefit (if the patient even has one). Patients sometimes have an unpleasant surprise when they go to pick up their discharge prescriptions and discover that the medications they received in the hospital are extremely expensive when they leave.

The short-term perspective is important for hospital psychiatrists. It is critical to be able to focus on the here-and-now and not to become overwhelmed with questions about how the patient might manage long-term problems such as the possible need for disability benefits or a move into assisted living. But the patients who cycle through the hospital do have to worry about these things. It can be overwhelming for patients whose problems started long before the hospital, and will extend long after they are discharged, to be rushed through the hospital process without being able to address some of those issues. During hospitalization, clinicians sometimes miss important details about how patients live their lives.

One way to address this focused perspective of the hospital is to emphasize to patients that we can only address a small part of their lives during the hospital stay. For patients who experienced hospital care decades ago, this will mean managing their expectations to emphasize the change from the past. Other ways of addressing the hospital perspective are to have activity groups that discuss longer-term life issues such as work and leisure. Good collaboration with pharmacists can help with thinking about how patients take medications in the real world and which medications are covered by their insurance.

Staffing a hospital service is more complicated than managing outpatient appointments. Hospital emergency services cannot control (or sometimes even predict) the volume of patients who cycle through at any given time. Emergency services likely will be either overstaffed, which can be expensive, or understaffed, which can lead to increased risk when managing acutely psychiatrically ill patients. Consultation services on medical and surgical units can also fluctuate in terms of need. Sometimes medical providers are easily able to manage the emotional or behavioral issues that arise in their patients, but sometimes they ask for more help or they have more patients with major psychiatric problems (such as suicide attempts). It is somewhat more predictable to staff an inpatient psychiatric service, but the length of stay of patients can vary to a large extent, which leads to variable levels of work.

An ideal hospitalist model would have a cadre of psychiatrists who could staff any of the possible areas as demand fluctuated. The overlap in skills between emergency consultation and medical and surgical consultation is extensive, and emergencies drive inpatient admissions. Here again, the advantage of a hospitalist model is that physicians can be deployed when they are needed. An inpatient unit that varies in size based on the availability of its physicians will be less efficient and run the risk of losing resources needed to run the other essential parts of the unit (such as nursing staff and activity therapy).

All states now have some kind of system of rules to protect patients’ rights, specifically around the circumstances in which patients can be kept on locked inpatient psychiatric units and given treatment against their will. Whether or not current psychiatrists appreciate the resulting processes, the states’ mental health legal procedures were enacted in response to decades of physician carelessness and what appeared to be patient abuses within long-stay hospitals. Physicians who practice in hospital settings should be aware of the rules that operate within their state regarding inpatient commitment. Some inpatient psychiatrists will need to testify before probate judges about their patients’ need for treatment. Court processes can be important to help with residency training, but it is ideal to have some specialized help or guidance from the court for psychiatrists to learn the correct methods of interaction in that setting. Legal issues are discussed in detail in Chapter 14, “Legal and Ethical Issues.”

Most insurance companies require a physician or social work evaluation to authorize an inpatient admission. The requirements to justify authorization can vary somewhat depending on the insurance company. Some are stricter about demanding that a patient articulate a clear plan and intent for suicide prior to authorizing hospitalization. Others are more willing to understand that sometimes a patient has failed outpatient management and needs a higher level of care. Hospitals must have a staff or system to manage the myriad insurance requirements to ensure that authorization is obtained and to make it clear to the psychiatrist what level of detail in documentation or telephone communication is necessary.

After the initial authorization, most insurance companies require hospitals to demonstrate that patients still meet criteria for continued stay. Again, some companies want hospitals to document that patients are still suicidal with a plan; they might deny continued authorization if a patient is still having suicidal thoughts but no active plan. Some companies move quickly toward a peer review process in which a physician contracted by the insurance company speaks to a hospital psychiatrist about why the patient is still in the hospital. Once again, it is critical for a system to be in place to help the hospital psychiatrists document in the most effective way possible and to facilitate the peer-to-peer review requests that occur.

Hospitals are tightly regulated places to work. The Joint Commission, the regulatory body that accredits most hospitals in the United States, has thousands of rules and regulations about what is allowed, what is required, and what is permissible to show that a hospital is meeting an acceptable standard of care. The rules not only are complex but also continue to change. The Joint Commission’s rules are enforced by the Centers for Medicare and Medicaid Services, which will withhold payment to hospitals for government insurance programs if hospitals fail to meet specific standards. Although the full management of regulatory requirements necessitates a specific person or an office within a hospital because of the complexity, hospital psychiatrists need to have a basic understanding of how the regulations affect patient care.