The task of evaluating psychiatric disability poses several ethics-related and practical challenges for psychiatrists, especially when they are responding to a request from a third party for a disability evaluation on their own patient. This study sought to evaluate the differences in how forensic and nonforensic psychiatrists approach and view evaluations for Social Security disability benefits. Thirty-two forensic and 75 nonforensic psychiatrists were surveyed on their practice patterns and perceptions of role, objectivity, and dual agency in the disability evaluation process. Significant differences were found between forensic and nonforensic psychiatrists' perceptions of the dual-agency conflict, beliefs about who should perform evaluations, and beliefs about the weight given to different opinions when decisions of whether to award disability benefits are made. A minority of respondents in both groups reported having identified a patient as disabled, despite believing otherwise. The implications of these findings are discussed.
The Social Security Administration (SSA) provides federal benefits for approximately 3.4 million adult Americans disabled by mental illnesses, representing a 260 percent increase from 2000 to 20081,2 and nearly one third of all SSA beneficiaries, the largest of any diagnostic category.1 The lost work productivity from psychiatric disability in general has been estimated at $150 billion yearly,3 with $44 to $51.5 billion attributed to depressive disorders alone.4,5 Because a large number of Social Security claims are denied on their first submission,6 only to be appealed and later litigated in court,7 it may be years before a final disability determination is made.
It is not surprising, therefore, that disability assessments of patients are the most common nontherapeutic evaluations requested of psychiatrists.8,9 The high number of evaluations performed, the substantial impact that determinations have on patients' lives, and the collective societal costs of psychiatric disability underscore the need for psychiatrists and other health professionals to provide objective and reliable data on those who seek benefits. Yet several issues of law and ethics arise for psychiatrists who perform disability evaluations pertaining to their own patients, including the conflict between serving as clinician and objective evaluator (dual agency, i.e., the tension between wanting to advocate for patients versus portraying their impairments objectively),10 the need to obtain informed consent,8 applying a statutory or regulatory definition of disability in the evaluation process,7,11 having to correspond with lawyers by whom the majority of SSA applicants are represented,6 and, occasionally, providing court testimony.7
Psychiatrists must also contend with the possible pressure from and resentment of patients who disagree with how their impairments have been represented.12 To strengthen their applications, some patients misrepresent their symptoms13 or their adherence and response to treatment. Further straining the therapeutic alliance is the concern about how benefits already awarded may affect treatment13–15 and the need to assess periodically whether benefits should be terminated.16
Notwithstanding the available resources to help psychiatrists navigate these concerns, the inconsistent correlation between symptom severity and functional capacity17,18 suggests that clinical experience alone does not provide adequate preparation for assessing disability. Forensic psychiatrists, because of their training and ethics framework,19 may be better prepared to recognize and handle the challenges that disability evaluations pose.7 Nevertheless, the SSA gives greater weight to the input of treating psychiatrists7,10 because they are able to provide a more detailed and longitudinal picture of the claimant's symptoms and impairments.20 Federal courts have consistently supported this preference21–24 and have held that a treating doctor's opinion may be rejected only for clear and convincing reasons.25
Despite the frequency of mental health-related disability determinations, research on psychiatrists' experiences in evaluating psychiatric disability is largely nonexistent. In reviews of forensic topics that are essential for a training curriculum for general psychiatry residents,26–28 only one recommended including workers' compensation.28 Workers' compensation as a training topic was the only disability-related topic that was identified in a survey of residency training directors in the United States and Canada, and even that was identified as the subject matter covered least often and deemed the least important of the subjects in forensic psychiatry.29 More recently, senior psychiatry residents were found to underappreciate dual-agency conflict in disability assessment, display low confidence in their ability to assess disability accurately, and identify a need for more training on evaluating disability.30
We sought to identify the differences between psychiatrists with and without forensic expertise regarding their experiences and beliefs about the disability evaluation process, including how they perceive their role, how much weight they believe is given (and should be given) to their opinions in SSA disability determinations, and to what extent they appreciate the dual-agency conflict. We hypothesized that forensic psychiatrists would have a greater understanding of the dual-agency conflict than would their nonforensic colleagues.
A survey questionnaire was developed to assess general and forensic adult psychiatrists' experiences and beliefs about performing Social Security disability evaluations. The survey (available on request) included 21 items with scaled or counted responses and a comments section. Survey content domains included the frequency with which disability evaluations were requested and completed, the practice patterns for conducting evaluations, behavior with regard to intentionally misrepresenting disability, beliefs regarding the weight given to opinions of evaluators, and perceptions of objectivity, role, and dual agency in the evaluation process.
In 2003, 200 general psychiatrists and 100 forensic psychiatrists were selected from the membership directories of the American Psychiatric Association (APA) and American Academy of Psychiatry and the Law (AAPL), respectively. We wanted to draw from a subject pool representative of AAPL and APA members, and an equivalent number of subjects with last names beginning with each letter in the alphabet were therefore sampled from each directory. International members and members-in-training (fellows, residents, and students) were excluded. Individuals with membership in both the AAPL and the APA were excluded from the APA sample group and replaced by the next non-AAPL APA member on the list. Subjects were mailed the survey, a return envelope, and a consent letter explaining the study. Study participation was voluntary and confidential; the return of a completed survey constituted consent. The University of Massachusetts Medical School Institutional Review Board approved the study.
Descriptive statistics examined responses to survey items. The chi-square and Mann-Whitney U nonparametric analyses were used for selected group comparisons. An α level of .05 was used to determine statistical significance.
Ninety-seven (32.6%) of the 300 eligible respondents agreed to participate: 65 (32.5%) in the APA group and 32 (32%) in the AAPL group. The majority of respondents (72% of the APA group and 62.5% of the AAPL group) had been in practice over 10 years. Of the APA respondents, 95.4 percent (n = 62) reported spending five percent or less of their time engaged in forensic-related work, whereas 78 percent of the AAPL respondents spent more than five percent of their time (χ2 = 56.43; df = 1; p = <.001). All the AAPL respondents identified themselves as forensic psychiatrists, and all the APA respondents identified themselves as nonforensic psychiatrists.
Regarding practice patterns, the APA respondents were significantly more likely than the AAPL respondents to have been asked to complete disability forms (83.1% (n = 54) versus 62.5% (n = 20); χ2 = 5.02; df = 1; p = .025). As shown in Table 1, of those asked to complete disability forms, the APA respondents were significantly more likely than the AAPL respondents to indicate that they agreed to complete the forms (p = .008) and, at a trend level, to report feeling pressured by patients to do so (p = .078). The respondents were asked whether, to help a patient, they had ever indicated on an SSA form that the patient was disabled, despite believing that the patient could work; 14 percent (n = 8) of the APA and 20 percent (n = 4) of the AAPL respondents reported that they had done so. No significant difference was found between the APA and AAPL respondents in the practice of obtaining informed consent for the disability evaluation and the frequency of completing evaluations over the preceding year.
Regarding role identification, 71.7 percent (n = 38) of the APA and 72.2 percent (n = 13) of the AAPL respondents stated that they complete evaluations because they identify it as part of their role as the treating clinician; 50.1 percent (n = 27) and 38.9 percent (n = 7) because as the treating clinician they know the patient better than would an independent evaluator; 35.8 percent (n = 19) and 33.3 percent (n = 6) because they want to help patients qualify for SSA benefits; 20.8 percent (n = 11) and 27.8 percent (n = 5) because they want to maintain the therapeutic alliance; and 20.8 percent (n = 11) and 11.1 percent (n = 2) for other reasons. (Respondents could choose more than one item.) When completing disability forms, 67.9 percent (n = 36) of the APA and 77.8 percent (n = 14) of the AAPL respondents identified themselves as patient advocates; 13.2 percent (n = 7) and 27.8 percent (n = 5) as forensic experts; and 17 percent (n = 9) and 11 percent (n = 2) as professionals employed by the state or other government. (Respondents could choose more than one item.)
Regarding beliefs about who should perform disability evaluations, the APA respondents were significantly more likely than the AAPL respondents to believe that only treating clinicians should perform disability evaluations (p = .008). However, responses did not differ significantly between the APA and AAPL members on whether it matters if independent examiners versus treating psychiatrists perform disability evaluations (p = .132) as long as they are done properly, whether only independent examiners should perform evaluations (p = .974), or whether it is more cost effective to have treating physicians perform evaluations (p = .154).
Regarding the weight the SSA gives to evaluator opinions, the APA respondents were significantly more likely than the AAPL respondents to believe that the SSA gives more weight to the opinions of independent examiners (p = .023). At a trend level, the AAPL members were more likely to believe that the SSA gives more weight to the opinions of treating clinicians (p = .068), and the APA respondents were more likely to believe that treating clinicians' opinions should carry the most weight in disability determinations (p = .085).
As shown in Table 2, the AAPL respondents were significantly more likely than the APA respondents to agree that the problem of dual agency negatively affects the disability determination process (p = .004) and that the physician-patient relationship is affected when treating physicians perform disability evaluations (p = .02).
When the APA and AAPL respondents were pooled, those who had indicated that a patient was disabled despite believing otherwise had completed fewer evaluations in the preceding year (χ2 = 7.24; df = 2; p = .027) and were more likely to feel that the physician-patient relationship is affected when treating clinicians perform disability evaluations (U = 228; z = −2.4; p = .017). At a trend level, these respondents were more likely to identify themselves as an advocate when completing disability forms (χ2 = 2.73; df = 2; p = .098), to agree to complete evaluations because they want to help patients get benefits (χ2 = 3.51; df = 1; p = .061), to believe that independent examiners give more objective opinions regarding disability (U = 260; z = −1.92; p = .054), and to believe that the SSA gives more weight to opinions of treating clinicians (U = 222.5; z = −1.7; p = .09). No relationship was found between the likelihood of identifying patients as disabled despite believing otherwise and the length of time practicing psychiatry (p = .798), percentage of practice involving forensic work (p = .306), frequency of agreeing to complete evaluations (p = .766), or the pressure felt to complete evaluations (p = .232).
To our knowledge, this is the first study to examine the beliefs and experiences of psychiatrists who perform Social Security disability evaluations. As hypothesized, we found that psychiatrists with forensic expertise were more likely than those without to identify the dual-agency conflict as negatively affecting the disability determination process and the impact on the treatment relationship of having clinicians perform evaluations on their own patients.
This study should be interpreted in the context of several limitations. The sample size and response rate limit the generalizability of our findings, since we cannot determine how representative our respondents were of general and forensic psychiatrists. Individuals with limited experience in performing disability evaluations may have been more likely to ignore the survey. As with all surveys, responses are subject to recall bias and individual interpretation of survey questions. Strong personal beliefs and frustrations with the disability evaluation process may also have motivated some to participate in our study or respond in a certain way. With these limitations in mind, this study has several potentially important findings.
First, psychiatrists who affiliate themselves with forensic psychiatry by virtue of their AAPL membership were more likely to agree that there is a conflict in evaluating disability in the patients whom one treats. This difference corresponds with two other findings: that the AAPL psychiatrists tended to feel less pressured to complete evaluation forms and were significantly less likely to agree to perform evaluations on their own patients. Although these data suggest that AAPL psychiatrists better recognize dual-agency concerns and make efforts to avoid such conflicts, they do not necessarily indicate that AAPL psychiatrists are more capable of managing them in the evaluation process. Most of the AAPL psychiatrists still responded that they felt pressured to complete forms “always” or “usually,” did not consistently obtain informed consent from their patients for the evaluation process, and primarily identified themselves as advocates when performing disability assessments. It seems, then, that many psychiatrists, regardless of their forensic expertise and recognition of the dual-agency conflict, struggle with the management of overlapping therapeutic and forensic roles when performing disability assessments.10
A substantial minority of the respondents in both groups reported having identified a patient as disabled despite believing that he could work. In wording this question, we specifically sought to identify instances in which the psychiatrist purposefully misrepresented a patient as disabled to help him rather than instances in which the psychiatrist may have erred on the side of endorsing disability when the evidence was equivocal. Because this question addresses a particularly sensitive concern in the disability determination process and psychiatrists may have been reluctant to report having engaged in this behavior, our finding may underestimate the number of respondents who have identified a patient as disabled contrary to their beliefs about the patient's actual functional capacity. That those psychiatrists who reported having misidentified disabled patients had also performed fewer evaluations over the past year is not surprising. The infrequency with which they perform disability evaluations may make them less familiar and comfortable with the primary importance of objectivity in this role. However, those who had misidentified a patient as disabled were more likely to agree that evaluating one's own patients affects the treatment relationship, tended to be more likely to see themselves as advocates, and tended to be more likely to agree that the SSA gives preferential weight to the opinions of independent examiners when making disability determinations. It is unclear why some psychiatrists, despite an awareness of dual-agency concerns, reported having intentionally misrepresented their patients' impairments. Perhaps they rationalize that the SSA will right their wrong after considering the input from an independent examiner or, seeing themselves as an advocate for their patients, will feel that they are ethically justified, given the inherent conflict in dual-agency assessments. Whatever the reason, this practice, which is of particular concern, likely reflects the difficulty psychiatrists have in performing a forensic task faithfully when it poses a risk to the therapeutic alliance. Further study of this topic is warranted.
Although the non-AAPL psychiatrists were more likely to believe that only treating clinicians should perform disability evaluations, they largely did not recognize that the SSA gives greater weight to the opinions of these clinicians. It is troubling that so many of the non-AAPL psychiatrists failed to appreciate the value and function of their clinical observations and opinions in disability determinations. Perhaps the substantial variability in the processing of SSA disability applications and the high rates of initial denial of claims and reapplications6,31 leaves treating psychiatrists confused about how their input is used. Indeed, state-to-state approval rates for all SSA disability claims vary by as much as 30 percent each year, calling into question the integrity and the fairness of the SSA disability system.32 Moreover, in 2000, 50 to 75 percent of initially denied SSA claims were subsequently awarded when subjected to judicial review.31 Such inconsistencies are likely to frustrate patients and psychiatrists alike and have been major considerations in the push to develop new approaches to determining disability.32,33
Approaches to disability assessments have common themes but may differ depending on the context in which they arise. For example, the depth of data that a treating psychiatrist may provide for an SSA disability determination often differs from the more comprehensive data collection and opinions that may be required for disputed SSA claims or private disability forensic reviews. Overall, our findings highlight the need for better education of general and forensic psychiatrists who perform (or might be asked to perform) psychiatric disability evaluations. The AAPL Practice Guideline for the Forensic Evaluation of Psychiatric Disability7 is a welcome resource to this end. Given the high prevalence of psychiatric disability and its associated costs, general psychiatry residents and forensic psychiatry fellows may further benefit from the development of a formal curriculum on this topic, and it should be regarded as an important area to cover in psychiatric continuing education.
We found that psychiatrists with forensic affiliations are more likely to identify a dual-agency conflict in evaluating their own patients for Social Security disability benefits. In general, treating psychiatrists struggle with managing their therapeutic and forensic roles, and a small minority misidentify patients as disabled despite believing otherwise. Psychiatrists without forensic affiliation underappreciate the weight of their opinions in the SSA disability determination process. These preliminary findings highlight the need for further research and better education on the practices and challenges in evaluating psychiatric disability and the weight given to different opinions by the SSA.
Disclosures of financial or other potential conflicts of interest: None.
- American Academy of Psychiatry and the Law