Barriers to the Implementation of Electroconvulsive Therapy (ECT): Results From a Nationwide Survey of ECT Practitioners

Electroconvulsive therapy (ECT) is an effective treatment for major depressive disorder and other psychiatric disorders. When treating major depressive disorder, first-line ECT can produce remission rates of 80%–90%. In regard to treatment-resistant depression, ECT remission rates are 50%–70% in clinical trials (1). This finding compares favorably with oral antidepressant remission rates of 13%–14% for treatment-resistant depression from the Sequenced Treatment Alternatives to Relieve Depression trial (2). In addition to efficacy for major depressive disorder, recent trials have confirmed that ECT is efficacious among individuals with schizophrenia that is resistant to clozapine (3) and among individuals with bipolar depression that is treatment resistant (4). With respect to population health, ECT has been shown to lead to lower rates of rehospitalization (5) and to mitigate suicidal ideation (6); in addition, ECT may be cost effective after just one or two failed oral antidepressants (7).

Although major depressive disorder is not the only indication for ECT, it is the most common one (8). However, growing evidence suggests that a small portion of patients with depression actually receive ECT. Using the MarketScan data set of privately insured individuals (representing 47 million Americans), researchers showed that approximately 0.25% of individuals with major depressive disorder received one or more ECT treatments in 2015 (9). Even among patients older than age 65, for whom ECT is particularly helpful (10), another recent analysis that used Medicare data showed that ECT was provided for only 0.41% of patients with major depressive disorder (11). The rate of ECT use has decreased substantially, especially among older patients, since the mid-1990s (12).

Expert opinion suggests that several factors might be limiting the implementation of ECT, including stigma, bureaucratic rules on reimbursement, high cost, and lack of adequately trained staff (13). However, no systematic investigation has been conducted regarding the relative contribution of different barriers to implementation. We sought to quantitatively examine the barriers to ECT implementation from the perspective of the provider via a nationwide survey. We also examined regional variation in ECT availability on the basis of the Medicare Provider Utilization and Payment Database.

Methods

Qualitative Interviews

To understand barriers to establishing and expanding an existing ECT service, we conducted semistructured qualitative interviews with U.S.-based ECT providers (N=17) who responded to solicitation via the e-mail Listserv of the International Society of ECT and Neurostimulation (ISEN). Interviews were recorded and transcribed verbatim. We analyzed the interviews by using qualitative thematic analysis procedures in which coding and subcoding categories were derived iteratively from transcribed data, integrating inductive and deductive approaches (14, 15). Two members of the research team (S.T.W., A.H.) separately coded the interviews by using NVivo, version 11, they met twice during the process to reach consensus on final themes.

Quantitative Survey Development and Deployment

Next, we created a quantitative survey based on themes identified by the qualitative analysis. We assessed provider perceptions as to the relative importance of barriers to expanding an existing ECT service or starting a new service in the same geographic setting. From qualitative analysis, 22 potential barriers to expanding an existing service or starting a new service were identified (for details, see the online supplement to this article). The comparative importance of these barriers as perceived by providers was assessed by asking participants to rank in order the most prominent five barriers to expanding an existing service and starting a new service. (For survey dissemination and eligibility, see the online supplement.) Comparing those who responded to the survey (N=192) with the population of ECT providers from the Medicare Provider Utilization and Payment Database (N=690), survey respondents were representative with respect to gender and geographic distribution (see online supplement).

Exploration of Regional Variation in ECT Services

We used the Medicare Provider Utilization and Payment Database to explore regional variation in the availability of ECT services and providers. We mapped the number of providers and ECT services in a given state to the total state population by using 2016 U.S. Census Bureau data. The numbers of ECT providers and ECT services per 1 million individuals were calculated and mapped. The project was approved by the Yale Institutional Review Board.

Results

Qualitative Results

Analysis of the qualitative component of the study showed that within hospitals, providers face barriers with hospital administration, including bureaucracy, stigma, and lack of understanding of the treatment and finances. In some cases, providers reported that hospital administrators do not value ECT compared with other procedural-based diagnostics or therapies (i.e., colonoscopy) because of relatively low reimbursement rates. Subsequent lack of administrative support leads to low resource allocation, including physical procedure space, support staff, and inadequate marketing of the treatment, resulting in invisibility both internal and external to the hospital. Barriers are also faced among support staff: collaboration of anesthesiologists is critical, but they often have competing interests and duties such that attending to the ECT service receives low priority; nursing staff often harbor negative perceptions of the treatment (similar to the general public) or are not adequately trained to provide ECT-related care; and even fellow psychiatric staff can have antiquated perceptions of the treatment. More successful programs have the support of a champion (one who is fully invested and committed to expanding access to the treatment) (16) within the administration and a commitment to accurate communication and education about ECT to hospital staff. When adequate space and staffing are provided, the service functions well, and the reputation and visibility of the service also improve among hospital staff and community providers.

Nationally, providers explained that the limited number of ECT centers and the small number of trained experts (resulting from the lack of training options) lead to inconsistent treatment approaches and protocols. Availability of treatment varies widely across the United States, and transportation is a significant barrier for those who want the treatment but do not live close to a providing facility. (Because they receive general anesthesia, ECT recipients are not permitted to drive themselves home following treatment.) These barriers pose a particular challenge for patients with low income. The shortage of treatment options (and the failure of some hospitals that offer the treatment to market it effectively) means that awareness is lacking about the treatment, which reinforces existing stigma among potential patients and their families (particularly among racial-ethnic minority groups) (17) often resulting from a history of unethical or inappropriate ECT use. Referral processes are not well developed and are sometimes disorganized. Reluctance to refer often occurs because of stigma among medical providers, ECT’s lack of prominence in treatment algorithms, and inappropriate referrals (i.e., ECT is seen by many psychiatric providers as a last resort or is used only when community psychiatrists do not know how to manage patients with challenging conditions). In some jurisdictions, legal barriers exist, and a court order is needed in cases in which a patient is catatonic and thus incapable of consenting (as opposed to the approval of a guardian or next of kin with similar clinical circumstances for nonpsychiatric procedures) (see online supplement).

Quantitative Results

Demographic characteristics of respondents.

Completed surveys were received from 192 physicians, representing 142 hospitals. Most respondents were male (N=155, 81%), and the mean±SD age was 53.9±13.2. Respondents reported having been in practice (since graduating residency) for a mean of 21.6±13.5 years, with a substantial portion (N=91, 47%) reporting more than 15 years of ECT experience. Approximately one-quarter (N=43, 22%) were international medical graduates.

Training of respondents.

The largest proportion of providers responded that they had received hands-on training for ECT (N=74, 39%), whereas slightly more than one-quarter reported having completed the ISEN training program (N=53, 28%) or a mini-fellowship (N=54, 28%). Just 7% (N=13) of respondents reported receiving their training as part of a geriatric fellowship, with fewer (N=11, 6%) reporting a formal, year-long ECT fellowship. Almost one-third of respondents (N=58, 30%) reported having graduated residency from one of just 12 institutions (see online supplement).

Characteristics of hospitals and ECT services represented.

Most respondents practiced at teaching hospitals (N=126, 66%) or in an urban setting (N=178, 93%). The mean number of psychiatrists in a given practice was 3.32±1.84. Roughly equal proportions of services used a Mecta (N=65, 46%) or a Thymatron (N=71, 50%) device, with 3% (N=4) reporting use of both. Most ECT services (N=111, 78%) had been in existence for more than 15 years. Just more than half of hospitals used a dedicated ECT space (N=83, 58%), whereas one-third (N=48, 34%) used a shared postanesthesia care unit (PACU) for administration of treatments. One-third (N=47, 33%) of hospitals represented in the survey performed 20 or fewer treatments in a given week, with just less than one-third (N=44, 31%) in the range of 21–40 treatments per week, and 35% (N=50) with greater than 40 treatments per week. For convenience, these ECT services were designated as small, medium, and large. A significant relationship was found between the size of the ECT service and the type of ECT space (dedicated space vs. PACU), and services with a dedicated ECT space were much more likely to provide a greater number of treatments (χ²=25.87, df=1, p<0.001) (Table 1).

Barriers to expanding existing service.

By rank-order scale scores, the most prominent provider-reported barriers to expanding an existing ECT service were (in decreasing order of importance) lack of physical space, stigma on the part of patients, transportation problems, lack of administrative support and bureaucratic issues, lack of adequately trained ECT practitioners and colleagues, potential side effects not acceptable to patients, lack of knowledge on part of potential referring providers, and poor reimbursement rates (Table 2). When stratifying by university-affiliated teaching versus community hospital, practitioners from teaching hospitals reported that lack of physical space was a more important barrier (1.93 vs. 0.98; t=8.07, df=141, p=0.005) (see online supplement). Lack of knowledge by potential referring providers was ranked as more of a barrier in community hospitals compared with teaching hospitals (1.40 vs. 0.86; t=2.14, df=141, p=0.033). Patient volume (i.e., large vs. small ECT service) did not affect reported barriers.

Barriers to starting a new service.

By rank-order scale scores, the most prominent provider-reported barriers to initiating a new ECT service were (in decreasing order of importance) lack of well-trained colleagues and ECT practitioners, lack of a champion, lack of physical space, lack of administrative support and bureaucratic issues, lack of well-trained support staff (nursing), stigma on the part of the patient, poor reimbursement rates, and transportation problems (Table 3). Compared with community hospitals, practitioners from teaching hospitals reported that lack of a champion was a more important factor (2.21 vs. 1.39; t=4.68, df=141, p=0.031). Practitioners from small services reported that poor reimbursement rates were more of a barrier compared with large services (1.20 vs. 0.60; t=7.24, df=141, p=0.007).

Regional variation in availability of and barriers to ECT.

Wide regional variation was seen in the availability of ECT. The mean number of ECT providers per 1 million individuals per state was 2.97±2.62, with a median of 2.23 and a range of 0 (Alaska and Idaho) to 12.8 (Maine). The U.S. census division with the highest per capita number of ECT providers was New England (6.4), and the lowest was the West South Central (1.1) (Figure 1) (see online supplement). No major regional variations were found in provider-reported barriers to initiating or expanding ECT services.

Discussion

To our knowledge, this study is the first attempt to quantitatively compare provider-perceived barriers with ECT service expansion or initiation and the first nationwide survey of ECT providers. Qualitative data showed at least 22 barriers to initiating or expanding an ECT service. With respect to expanding an existing service, the most prominent provider-reported barriers were lack of physical space, stigma on the part of patients, and transportation issues. With respect to initiating a new service, the most prominent provider-reported barriers were lack of well-trained colleagues and ECT practitioners, lack of a champion, and lack of physical space. We also found great geographical variation in the availability of ECT providers.

Lack of physical space was identified as a prominent barrier for both initiating new ECT services and expanding existing ones. From qualitative interviews, this barrier seemed tied to low reimbursement rates of ECT relative to other procedure-based diagnostics and therapies (i.e., colonoscopies). The narrative that emerged from these interviews was that, from an administrative perspective, ECT is often the lowest on the “procedural totem pole.” If another higher-revenue clinical service needs additional space, ECT services might be displaced.

Expectedly, our data show that having a dedicated space in which ECT was performed was significantly related to the size of the ECT service (Table 1). If ECT is performed in space borrowed from a surgical recovery area (i.e., a PACU), all ECT patients must be ready for discharge by the time that the first wave of surgeries starts to finish (approximately 10 a.m.). In contrast, services with dedicated space can schedule patients into the afternoon, accommodating a greater volume of patients per day and increasing revenue. Therefore, advocating for dedicated ECT space requires forward-thinking financial projections that include potential cost savings by, for example, offsetting rates of readmission and duration of inpatient hospital stay if outpatient services can be accommodated. It should be noted, however, that this relationship could be bidirectional; hospital administrators may not be willing to invest capital in smaller ECT services.

Two potential remedies for this barrier include enhancing the reimbursement rate for ECT (which can be done on a local basis with private insurers), to enhance the value of ECT in the eyes of administrators, and adjusting Centers for Medicare and Medicaid Services rules that do not allow for reimbursement of ECT in ambulatory surgical centers (current regulations mandate that ECT be performed in a hospital). The latter policy change would create more of a “buyer’s market” for ECT practitioners and might be cost saving from the perspective of the government (facility fees for hospitals are generally higher than ambulatory surgical centers). Unfortunately, recent efforts to appeal to federal policy makers in this respect have not been successful (see online supplement).

Stigma on the part of patients was another main barrier to the expansion of existing ECT services. Undoubtedly, this stigma is rooted in past practices and inaccurate media representations of ECT. Data have shown that the majority of media representations of ECT are negative or inaccurate. In American media, most movies or TV shows depict ECT being performed without anesthesia or done for the purposes of behavioral control or torture (neither of which are accurate in contemporary psychiatric practice) (18, 19). Engagement of stakeholders as well as media producers may be important in further reducing stigma toward ECT. Research has shown that psychoeducation about ECT can improve perceptions and knowledge (20, 21); hence, proper education during referral may be a critical aspect to assisting prospective patients to understand the potential benefits and risks of treatment.

Another aspect of stigma likely comes from the real possibility of cognitive side effects from ECT. Fortunately, most patients do not experience substantial cognitive side effects (22, 23); however, some anecdotal reports exist of patients losing large periods of memory (24), a worrisome prospect for a potential ECT patient. The refinement of ECT techniques (i.e., right unilateral ultrabrief pulse) (25) has further reduced the probability of cognitive burden, with additional promising techniques on the horizon (26–28).

Transportation was the third most significant problem cited by providers in expanding an existing ECT service. As with any procedure performed under general anesthesia, patients cannot safely drive themselves home from ECT treatment. This restriction can be difficult if patients live far away from a treatment center or if they have few family and community supports. Indeed, at one site, approximately one-quarter of ECT patients referred were not able to be accepted because of transportation (29). Creative solutions involving multiple stakeholders, particularly third-party payers, may be necessary to overcome this barrier. Given the logistical requirements of ECT, the importance of a champion was deemed critical to starting a new service. Initiating a new service requires recruitment and coordination of anesthesia and nursing support (sometimes in the face of stigma and misunderstanding), administrative approval, purchasing necessary equipment, and marshaling space resources.

Out of more than 200 accredited psychiatry residency programs, a relatively small number (N=12) produced almost one-third of practicing ECT doctors who responded to the survey. In addition, this issue was ranked as the number one barrier for starting a new ECT service. These findings suggest that increased investment into education and training requirements could substantially increase the number of adequately trained ECT practitioners. Currently, psychiatry residency graduation criteria from the Accreditation Council for Graduate Medical Education require that residents be exposed to some ECT didactic. Hence, a large proportion of graduating psychiatrists will have never observed an ECT procedure (only one in 10 hospitals offers ECT) (12). The 12 programs that produced a large proportion of practicing ECT psychiatrists in our survey are generally known within the ECT community for rigorous and robust ECT programs; furthermore, many have additional local training requirements (i.e., a 1-month mandatory rotation with ECT service) to ensure robust training of residents with respect to ECT. Greater investment in ECT training opportunities will not only expand the potential of new providers but will also minimize provider stigma, enhance comfort in providing psychoeducation to patients, and ultimately increase the likelihood of successful referral for treatment.

Several limitations of this study require comment. First, the survey response rate was poor (28%). Although survey respondents did not differ significantly from nonrespondents in terms of age and geographical region, respondents could differ in other important ways. Sampling bias likely existed, limiting the generalizability of our findings. Second, the Medicare Provider Utilization and Payment Database includes providers only if they render a given clinical service to 10 or more Medicare patients in a given year. Hence, our analysis may have missed some small ECT services or providers who do not accept Medicare. Third, our analysis took into account only barriers to treatment that were perceived by providers to be important. The actual comparative importance of these barriers may be different. Furthermore, other key stakeholders (patients and hospital administrators) may have different perspectives.

Conclusions

In this study, we have presented data from the first nationwide survey of ECT providers demonstrating perceived barriers to initiating and expanding ECT services. These barriers include lack of well-trained colleagues and ECT practitioners, lack of a champion, lack of physical space, stigma on the part of patients, and transportation issues. Coordinated efforts on overcoming these barriers may allow the treatment to be more broadly implemented.