Malingering in the Psychiatric Emergency Department: Prevalence, Predictors, and Outcomes

Malingering is the intentional fabrication of symptoms that do not exist or the exaggeration of genuine symptoms for the purpose of obtaining an external gain (1). Common gains include obtaining goods or services, such as accommodations in the hospital or disability benefits (2), and avoiding responsibility, such as legal action or attendance at work (3).

Malingering is a deceptive behavior (4) employed episodically and opportunistically. It can involve medical or psychological symptoms and can occur in the presence or absence of psychopathology. Malingering can be exploitative and for profit, although in certain circumstances it can be a more socially acceptable response to adversity. Malingering is not a psychiatric disorder and is considered a nonpathologic condition that may be a focus of clinical attention (1). There are no treatments indicated for malingering and no known role for intensive treatment settings, such as inpatient hospitalization.

The prevalence of malingering varies depending on the context and incentives sought (5, 6). Malingering in the psychiatric emergency setting has not been widely studied. In a 1996 study in a psychiatric emergency service, psychiatric residents reported that they strongly or definitely suspected 13% of patients to be malingering (7). In several studies, at least 10% of psychiatric inpatients with suicidality anonymously admitted to having lied or purposely exaggerated suicidal ideation to gain admission or during the hospitalization (8, 9). Over 4% of emergency visits may be from patients who fabricate symptoms to obtain medication (10). In a survey of neuropsychologists, malingering or symptom exaggeration was estimated in 8% of medical-psychiatric cases not involving litigation or compensation (11). Rates have been found to differ in other countries (12). Data on the prevalence of malingering need to be updated, and a description of the incentives and characteristics of malingered presentations in the psychiatric emergency setting, especially in a high-volume urban hospital, is also needed.

There are multiple benefits to better understanding malingering, including financial, safety, and care quality. Malingering among adults claiming Social Security disability benefits for mental disorders was estimated to cost more than $20 billion in 2011 (13). Malingering can distract clinicians from genuine symptoms, leading to missed diagnoses, and poses risk of future misdiagnosis should symptoms later be viewed as less credible. Unnecessary testing and treatments can lead to iatrogenic harm and reinforcement of deceptive behavior.

In the emergency department, malingering utilizes resources and is associated with high-risk situations, such as the staff-intensive discharge of a malingering patient who makes provocative statements or displays threatening behaviors (14). The evaluation of malingering is associated with escalation of the behavior (15) and aggression (2). Although countertransference can be useful to understand these situations, strong negative reactions from clinicians can further escalate situations (16). Identification of patient characteristics associated with a strong suspicion of malingering and guidance on the subsequent management of such patients, especially the decision to hold patients in the emergency setting for further observation, would be useful to improve understanding of physician decision making and resource allocation. Because malingering can be an adaptive response to adverse circumstances, understanding this behavior in the emergency setting may identify unmet needs.

The assessment of patients who report suicidal or violent ideas ideally relies on patient cooperation (17), which is lacking in the case of malingering (18). Provocative or goal-driven responses are barriers to assessment (19). When malingering is present, the clinician’s interview, documentation, and disposition planning are significantly affected (14, 20). Identifying and managing malingering is necessary for appropriate diagnosis and treatment and has recently been described as part of a residency training curriculum (21). Despite its role in the comprehensive assessment of patients, malingering is often not recorded in the medical record for various reasons, including associated lack of reimbursement and reluctance to document malingering for high-risk patients (8); even with accurate assessment, there is concern for legal liability (2).

Given the need to better understand malingering in the emergency setting and considering the complicated issues regarding its documentation, a purpose of our study was to conduct a systematic examination of malingering by administering questionnaires to psychiatrists after every patient encounter. Specific aims were to estimate the prevalence of malingering, determine whether rates increase in the second half of the month when financial needs may be higher, and identify the most common malingered presentations and incentives for malingering. We aimed to determine whether certain characteristics are associated with a high suspicion of malingering. In addition, we investigated how clinicians’ suspicion of malingering affects disposition decisions. We hypothesized that patients suspected of malingering would be more likely to be discharged from the emergency department, and we sought factors predictive of cases in which they instead were admitted to the hospital.

Methods

This study was evaluated by the Program for the Protection of Human Subjects at Mount Sinai Icahn School of Medicine and exempted from institutional review board review. The study took place at Mount Sinai Beth Israel, an academic medical center in New York City, in the comprehensive program in emergency psychiatry (CPEP). The CPEP is a locked psychiatric evaluation unit separate from the medical emergency department, staffed 24/7 by an attending psychiatrist, with psychiatric resident, social work, and ancillary staff support. Patients arriving in the CPEP were triaged as medically stable. The study sample did not include patients in the medical emergency department who had psychiatric consultation unless those patients were transferred to the CPEP.

There were 14 attending psychiatrists working shifts in the CPEP during the study period. These physicians were asked to complete one questionnaire for every patient evaluated in the CPEP throughout the month of August 2017, with instructions regarding patient handoff to ensure that only one form was completed for each patient.

We operationalized the definition of malingering to exclude defensiveness (the minimization of symptoms in effort to appear well) and nonmalingering provocative behavior and to specify both intentionality (to exclude somatoform and dissociative symptoms) and secondary gain (to exclude factitious disorder). Questionnaires asked the physician to “circle the extent that you suspect the patient is malingering (intentionally producing false symptoms or exaggerating symptoms, with secondary gain).” Possible responses were as follows: not suspected, slightly, moderately, strongly, or definitely. We used the scale described in the 1994 study by Yates and colleagues for comparison (7), which reflected malingering suspicion as a matter of degree (4).

Questionnaires were completed after comprehensive assessments, which involve subjective, objective, and collateral information. Assessments were made by attending psychiatrists, who have the most experience evaluating psychopathology and medical illness and using clinical judgment to discern primary and secondary gain (16). Although psychological testing for malingering could be helpful (22), it was not used because it may be impractical in the emergency setting (23) and likely to interrupt care and diminish the response rate.

Physicians noted patient gender, age, symptoms thought to be malingered, associated external gains, and initial disposition. The three disposition options were as follows: admit to inpatient psychiatry, hold for further assessment (which could include admission to an extended observation bed in the CPEP), or discharge without holding. During the study period, there was an approximately 4.5-day period when extended observation beds were unavailable, and physicians were asked to indicate their disposition as if this option were available. Questions about ethnicity and other demographic information were not asked in questionnaires to reduce the number of questions in the survey, and review of medical records was not part of the study.

During the study period, there were 429 patient encounters in the CPEP, and 414 questionnaires were returned. Nine of the 414 questionnaires were excluded because no malingering rating was indicated. The remaining 405 questionnaires were included in the analyses—a 94% response rate. One questionnaire noted that a patient evaluated in the CPEP was then admitted to a medical floor with psychiatric one-on-one observation. This case was coded as a psychiatric admission because this was the intention of the physician had the patient not developed an acute medical issue.

Data were analyzed with IBM SPSS, version 25. Initial analyses used chi-square tests, Fisher’s exact tests, analyses of variance, t tests, Mann-Whitney U tests, or Kruskal-Wallis tests, with two-tailed values of significance. Significant findings (p≤.05) were entered into logistic regression models to determine significant independent predictors of the variables studied.

Results

Malingering Frequency

Physicians reported suspicion of malingering for 33% of patients; 20% of all patients were strongly or definitely suspected of malingering. No differences were found between the first and second halves of the month regarding frequency of any suspected malingering, strongly or definitely suspected malingering, or distribution per degree of suspected malingering. During the first half of the month (N=190 patient encounters), 37% of patients (N=71) were suspected of malingering, with 22% (N=41) overall strongly or definitely suspected, compared with 29% (N=63) and 18% (N=39), respectively, in the second half (N=215 patient encounters).

Characteristics of Patients and Suspicion of Malingering

Compared with patients who were not suspected of malingering, those suspected were more likely to be male and had significantly different disposition outcomes (Table 1). No differences by age were noted.

Effect of Malingering Suspicion on Patient Disposition

Patients not suspected of malingering had an admission rate of 48%. As suspicion increased, the admission rate decreased. Patients who were strongly or definitely suspected of malingering had an admission rate of 4%. Discharge rates (excluding patients who were held) correspondingly increased with malingering suspicion.

Multinomial logistic regression was used to determine how each degree of malingering suspicion affected disposition outcome (Table 2). Compared with patients not suspected, those slightly suspected of malingering had 3.65 times greater odds of being held for further observation rather than being discharged outright and greater odds of being either held or discharged than of being admitted to the hospital. Patients suspected strongly or definitely of malingering had the lowest odds of being admitted to the hospital, and both had 25-times greater odds of being held rather than admitted, compared with those not suspected of malingering. For patients strongly suspected of malingering, the odds of being discharged rather than admitted were 16-times greater, compared with those not suspected. For those definitely suspected of malingering, the odds of being discharged rather than admitted were over 33-times greater, compared with those not suspected.

Most Frequent Malingered Presentations and External Gains

Patients suspected of malingering were grouped to distinguish those with slight or moderate suspicion from those with strong or definite suspicion, based on disposition trends and to facilitate comparison with prior work (7). Among patients with strong or definite suspicion of malingering, the most frequent symptoms assessed to be malingered were suicidal ideation (58%) and depression (39%); for 44%, it was assessed that the malingering involved multiple symptoms (Table 3). The most frequent secondary gains among patients with strong or definite suspicion of malingering were hospital admission (54%) and staying in the CPEP (35%); 25% were assessed to be seeking more than one secondary gain.

Characteristics and Predictors of High Malingering Suspicion

Patients for whom there was strong or definite suspicion of malingering, compared with those for whom there was slight or moderate suspicion, were assessed to more frequently present with malingered suicidal ideation (58% versus 38%) and to malinger a greater number of symptoms (1.58 versus 1.25) (Table 3).

To determine whether either of these factors was an independent predictor of high malingering suspicion, ordinal logistic regression was used. Degree of malingering suspicion (slight through definite) was the outcome variable; the presence of malingered suicidal ideation and the number of malingered symptoms were entered as predictor variables. Each increase in the number of malingered symptoms was independently predictive of a higher degree of malingering suspicion (OR=2.28, 95% confidence interval [CI]=1.33–3.93, p=.003). Malingered suicidal ideation was not independently predictive of a higher degree of malingering suspicion.

Characteristics Associated With Disposition Among Suspected Malingerers

For patients suspected of malingering, four items were significantly different according to disposition decision: degree of suspected malingering, the presence of malingered suicidal ideation, and the external gains of seeking hospital admission and social work or housing services (Table 4).

Thirteen patients suspected of malingering were admitted to the hospital. Each of the 13 patients was assessed to be engaged in malingering of suicidal ideation or to seek admission, and 11 of the 13 patients (85%) were assessed to be engaging in both. There were 131 patients suspected of malingering with recorded symptoms and dispositions. Among the 66 patients who were suspected of malingering of suicidal ideation, 18% (N=12) were admitted, 32% (N=21) were held, and 50% (N=33) were discharged. Among the other 65 patients who were suspected of malingering but not of malingering of suicidal ideation, 2% (N=1) were admitted, 25% (N=16) were held, and 74% (N=48) were discharged. There were 110 patients suspected of malingering with secondary gains and dispositions recorded. Among the 54 patients thought to be seeking hospital admission, 22% (N=12) were admitted, 24% (N=13) were held, and 54% (N=29) were discharged. Among the other 56 patients not seeking hospital admission, 2% (N=1) were admitted, 30% (N=17) were held, and 68% (N=38) were discharged.

Predictors of Admission Among Patients Suspected of Malingering

Although only 10% of patients suspected of malingering were admitted to the hospital, we entered significant characteristics regarding disposition from the unadjusted analyses into a logistic regression model to determine whether there were independent predictors of admitting a patient who was suspected of malingering (Table 5). This analysis described the odds of admitting versus discharging, because these represent definitive dispositions; however, all three disposition groups were included in the multinomial regression model. Malingering of suicidal ideation, seeking admission, and seeking social work/housing services were each independently predictive of admission, compared with discharge.

Adjusted Role of Malingering Suspicion in Disposition Decision

The above regression model indicated that for patients suspected of malingering, when the analysis adjusted for the presence of malingered suicidal ideation and secondary gains of seeking admission and social work or housing services, each increase in level of suspicion of malingering was associated with increased odds of being discharged compared with being held (OR=1.71, CI=1.05–2.79, p=.030), being held compared with being admitted (OR=3.08, CI=1.17–8.07, p=.022), and being discharged compared with being admitted (OR=5.27, CI=2.01–13.85, p=.001).

Discussion

To our knowledge, this is the largest investigation of malingering in the psychiatric emergency department to date. Our findings describe the high frequency of suspected malingering, especially of suicidal ideation and for the gain of hospital admission, and highlight the clinical relevance of malingering suspicion for patient disposition.

Patients with the highest degree of suspicion of malingering were discharged rather than held for observation, which demonstrates the certainty of the physicians’ assessments and the lack of significant genuine symptoms necessitating further care in the hospital or unclear symptoms requiring further workup. Among patients who were strongly or definitely suspected of malingering, 71% were discharged (14% of all 405 encounters).

The findings suggest that malingering may be more prevalent in the emergency setting than previously reported. In the 1994 study by Yates and colleagues (7), 13% of 227 psychiatric emergency patients were strongly or definitely suspected of malingering by residents, although there was suspicion for 42% of patients overall. These differences may result from population and study variations, but they may also result from our use of attending physicians to determine malingering, who may assess with more sensitivity than trainees.

The finding that a high suspicion of malingering was predicted by the number of symptoms thought to be malingered may suggest that the most strongly suspected malingerers in this sample were more obvious in “overacting” the part, a malingering technique that has been described elsewhere (2, 24). Patients for whom the suspicion of malingering was only mild or moderate may be malingering less, may conversely be more skilled and evade high suspicion, or may represent a group of patients for whom malingering is less certain. Given these issues, disposition analyses included patients with any suspicion of malingering rather than any particular subgroup.

Suicidal ideation may be the most frequently malingered symptom because of its subjectivity and lethality, as well as the patient’s anticipation of the clinician’s reluctance to discount such a self-report. Although most patients suspected of malingering were discharged from the emergency department (and we hypothesize that many who were held were later discharged), malingering of suicidal ideation appears relatively effective: among patients for whom there was any suspicion of malingering, those suspected of malingering of suicidal ideation had 22-times greater odds of being admitted than discharged, compared with those not suspected of malingering of suicidal ideation, when the analysis adjusted for degree of suspicion and significant secondary motivations.

The most common secondary gains were to be admitted to the hospital and to stay in the emergency room. Neither particular secondary gains nor the number of secondary gains were associated with the degree of suspicion of malingering, although seeking admission and seeking social work or housing services were each associated with being admitted to the hospital. Social needs may be common among patients who malinger in this setting, and hospitalization may facilitate service coordination. Encouraging the use of community resources has been shown to reduce emergency room visits (25), and further research may determine which resources would be most helpful to reduce the number of patient encounters involving feigned psychiatric symptoms.

It has been shown that in low-income households, resources are exhausted at the end of the month, which may result in acute medical needs (26). Although the study period was limited, our findings suggest that the incidence of malingered psychiatric complaints may not be related to timing in the month.

Geographic variability in factors affecting social needs, such as substance use disorders (27) and homelessness (28), as well as the accessibility of services to meet such needs, may provide a context for our findings. It is not known whether malingering is more prevalent in winter months. This hospital’s urban location within walking distance to other hospitals may facilitate peregrinating psychiatric patients, among whom malingering is common (29).

This study was designed to investigate malingering in the context of routine clinical care in the emergency setting, with emphasis on promoting the survey response rate, minimizing disruption of patient care, and preserving confidentiality of physicians and patients. The high response rate suggests minimized reporting bias, and the large physician participant group increases generalizability. The design, however, did not facilitate gathering additional demographic, racial-ethnic, diagnostic, or longitudinal information about each patient and did not permit comparison of data across physicians or shifts. Although malingering was assessed by attending psychiatrists after a comprehensive assessment, there may be variability in both the ways individual clinicians determined malingering and in their final determinations. It is important to note that results of clinical assessments can differ from results of those obtained via psychological testing and patient report (6, 8). This study did not identify patients who presented repeatedly, which is relevant because experience with known procedures plays a role in developing a malingering strategy (30) and because high utilizers of psychiatric emergency services have been described to have distinctive characteristics (31). Information regarding presenting symptoms was not collected for any patient, and it is not known whether particular symptoms were more likely to be suspected as malingered or whether co-occurring psychiatric symptoms were present in addition to malingered symptoms. Prior research suggests that particular diagnoses may be associated with malingering (32). More comprehensive future studies will further advance understanding of the nature and purposes of malingering.

Conclusions

Malingering is not a rare clinical phenomenon (33), and our findings show that suspicion of malingering is common in the psychiatric emergency setting. The prevalence and perceptions of this behavior have implications for both patients and physicians regarding the treatment relationship. If a patient was suspected of malingering, discharge was likely, although factors such as seeking social work or housing services, seeking hospital admission, or feigning suicidal ideation each increased the odds of hospitalization. Malingering for community services may indicate unmet social needs, and future work may identify areas for intervention.