As mental illness remains prevalent in the United States, the use of clinical psychotherapy services and psychotropic medication continues to grow (Marcus and Olfson, 2010; United States Department of Health and Human Services, 2009). Although many Americans take psychotropic medication without being in psychotherapy, a significant proportion of patients engage in both (Marcus and Olfson, 2010). Concurrent treatment has proven to be most effective (Riba and Balon, 2008), particularly for major depressive disorder (MDD) (Cuijpers et al., 2014; Keller et al., 2000). Combined treatment can be provided in one of two manners: integrated treatment, in which the psychotherapist also prescribes; and split treatment, in which two different professionals provide each therapeutic modality. Split treatment has grown explosively since the 1980s and has become a standard structure of treatment for millions of Americans (VandenBos and Williams, 2000).
The standard recommendation for split care is that communication and collaboration take place between the two professionals caring for the same individual (American Psychiatric Association, 2010; Goin, 2001; Gutheil and Simon, 2003; Riba and Balon, 2008). However, before 2010, only one study had investigated whether such communication actually takes place. Hansen-Grant and Riba’s survey (Hansen-Grant and Riba, 1995) quantified communication between psychiatric residents prescribing medication and the institutional psychotherapists. They reported infrequent, poorly documented communication, and the absence of recorded patient consent for communication between providers. Avena and Kalman (2010) published a survey of practicing psychotherapists that assessed the occurrence of communication between the psychotherapists and prescribers in split treatment. Kalman et al. (2012) published a similar survey of practicing psychiatrists. In these recent studies, split care was found to be quite common: 36% of psychotherapy patients and 41% of psychotropic medication patients were receiving psychotherapy and medication from two different professionals. The surveys concurred in reporting that for nearly 1 in 4 (23% and 24%, respectively) patients in split treatment for 6 months or longer, no communication had taken place between the involved professionals. Communication between providers is clearly a complicated matter that is lacking consistency.
Missing from the literature, however, are the perspectives of patients themselves and the impact of clinician communication on patient outcome. By soliciting patient reports of split treatment collaboration and patient satisfaction, this study sought to address the following research gaps: 1) replicating previous studies that examined collaboration practices between providers (Avena and Kalman, 2010; Kalman et al., 2012), but from the perspective of the patients; 2) assessing the relationship between degree of provider collaboration and patient satisfaction. Survey methodology was used to maximize the ease with which a broad population of split-care patients could be reached.
English-speaking individuals 18 years of age or older engaged in split treatment completed a survey regarding their experiences in this arrangement. The survey was distributed to participants, dubbed “workers” on Mechanical Turk, via a link to Survey Monkey (Appendix 1). Amazon’s Mechanical Turk crowd-sourcing software application has frequently been used for behavioral research (Mason and Suri, 2012).
Five hundred and two usable responses (batches of 212 and 290) were generated within 24 hours of posting on Mechanical Turk. Characteristics of the respondents are shown in Table 1. Two items were added to the survey for the second batch. Ages of respondents ranged from 18 to 64 with an average age of 30.8 (SD = 9.5).
Of the 502 respondents, 396 (78.9%) were in psychotherapy with a psychologist, 66 (13.1%) with a social worker, and the remaining 40 (8%) with a different therapist (pastoral counselor, nurse, other). Two hundred and thirty-six (236) individuals (47.0%) received their psychotropic prescriptions from a primary care physician, 259 (51.6%) from a psychiatrist, and the remaining 7 (1.4%) from another prescriber (nurse practitioner, obstetrician, or neurologist). Only the second batch (n = 290) of participants responded to a question about the sequence of their split-care treatments: 98 (33.8%) subjects first started medication, 115 (39.7%) began psychotherapy first, and 77 (26.5%) reported beginning both treatments at about the same time. Reasons for seeking treatment are listed in Table 2. Participants were not restricted to one answer. All participants provided informed consent after a detailed description of the study.
Participants completed a 23-item survey developed by the researchers for this study. In addition to demographic questions, survey items included questions regarding the communication practices between providers, the professional relationship between providers, and subsequent comfort levels and satisfaction of participants in light of their split-care arrangement. To increase validity of this Mechanical Turk sample, participants were asked, “Are you seeing a psychiatrist/other medical professional for psychiatric medication and a psychotherapist (i.e., psychologist, social worker, counselor, etc.) for psychotherapy?” Respondents who answered “No” to this item were excluded, although some then correctly identified themselves as indeed participating in split care; their data were then included.
Participants included split-care patients completing online research via Mechanical Turk (MTurk). The study was advertised as a “Short survey on your psychotherapy and psychiatric medication treatments (completion time: approximately 5 minutes).” Eligibility criteria (age 18 or older; English as a primary language; participating in a split-care arrangement) were included with the study description, followed by an informed consent page, involving a complete description of the study procedures, IRB approval, and a statement that completion of the survey would indicate participants’ consent. Recruitment for the survey was posted in two batches. Two items were added to the second bath such that not all items were offered to all participants. Those whose responses were deemed valid were paid $1.00, a fee that based on the time required for the task (<5 minutes) is higher than most fees paid to workers through Mechanical Turk.
Given the exploratory nature of this study, data analysis consisted of descriptive statistics of each item, as well as two independent samples, two-tailed t-tests, and three chi-squares. These analyses enabled the researchers to identify significant differences in satisfaction and comfort level between groups based on whether or not their providers had been in communication. The study was approved by the Institutional Review Board of Fordham University.
The current exploratory study on the patient’s experience in split-care treatment produced the following results. When asked about the frequency of communication between their psychotherapist and their medication prescriber, 100 individuals (20%) reported that communication had taken place more than twice; 140 (28%) reported that their clinicians had spoken once or twice; 101 (20%) said no communication had taken place; and 161 (32%) did not know whether or not their two providers had ever spoken. Considering only those who knew whether or not communication had taken place (341 individuals), there had been no reported communication for 101 (29.6%) of the respondents. Subjects were asked about the impact of provider communication on their satisfaction with the split-care arrangement, yielding the following results: Out of the 240 subjects whose providers had communicated, 147 (61.3%) reported increased satisfaction, whereas out of the 101 subjects whose providers had not communicated, only 11 (10.9%) reported an increase (χ2 = 73.41, df = 2, p < 0.001).
We also inquired whether or not either the psychotherapists or the prescribers had ever obtained subjects’ permission to communicate with one another. Regarding psychotherapists, 297 (59%) respondents said their therapist had requested a release of information, 126 (25%) said no they had not, and 79 (16%) subjects could not recall. With respect to prescribers, 269 (53.6%) said their prescriber had requested a release of information, 153 (30.5%) said no, and 80 (15.9%) could not recall. Excluding those who could not recall if their provider had asked for permission, 126 (30%) of 423 respondents and 153 (36%) of 422 reported never being asked for permission by their psychotherapist and prescriber, respectively.
Subjects were asked whether their two professionals had known each other before treatment and whether this affected their comfort with the split-care arrangement. Of the respondents, 186 (37%) participants responded yes, their providers had known each other 216 (43%) responded no, and 100 (20%) did not know. Out of those participants who were aware of whether or not their professionals knew each other, 239 (59.5%) subjects said their answer to this had no impact on their comfort with the treatment arrangement; for 134 (33.3%) their comfort increased; and for 29 (7.2%) their comfort decreased. However, for participants whose professionals had not communicated, just 16.2% experienced an increase in comfort, compared with 53.2% of respondents whose two professionals had been in communication (χ2 = 64.67, df = 2, p < 0.001).
One hundred and twenty-six (25%) participants indicated that their clinicians were from the same insurance network, 68 (13.5%) indicated that their clinicians were from the same institution, and 82 (16.3%) indicated that their clinicians were from the same practice group. Two hundred and twenty-seven (45.1%) individuals endorsed none of these relationships. Three hundred and thirty four (66.4%) respondents stated that their comfort with the split-care arrangement was unaffected by their answer to the above item, 145 (28.8%) reported an increase in comfort, and 24 (4.8%) reported diminished comfort. For participants whose professionals had not communicated with each other, just 9.5% experienced an increase in comfort compared with 35.6% of respondents whose two professionals had been in communication; in this way, participants whose providers communicated were significantly more comfortable with their treatment than participants whose providers had not communicated (χ2 = 34.87, df = 2, p < 0.001).
Subjects rated on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree) whether or not they felt that their psychotherapist and prescriber worked together during their treatment. “Agree” (value = 4) was the most frequent response (244 out of 501 participants, or 48.7%) with an average response of 3.51 ± 1.02 for the entire sample. Responses differed significantly between participants whose providers had communicated and participants whose providers had not (t = 14.5, df = 338, p < 0.001). The average score for the 101 respondents whose two providers had not had any contact was 2.58 ± 1.07. For the 239 respondents who reported communication between their providers the average score was 4.04 ± 0.73.
The second batch of respondents (n = 290) also answered an item assessing their satisfaction with their split-care arrangement. On a 5-point Likert scale (1 = very dissatisfied, 5 = very satisfied), 237 individuals endorsed being either satisfied or very satisfied (M = 3.88 ± 0.77). When sorted by provider communication (excluding those who did not know if communication had occurred), the difference in satisfaction between the two groups was not statistically significant (t = 1.86, df = 110, p = 0.065). The average score for satisfaction in the no communication group (n = 65) was 3.77 ± 0.86 and for those who reported contact (n = 144), the average score was 4.0 ± 0.76.
This study presents the results of a survey of 502 patients engaged in a split-care arrangement with a psychotherapist and a separate prescriber of psychotropic medication. Focusing on the question of communication between professionals working in split-care arrangements, this study supports the findings of previous psychotherapist and psychopharmacologist surveys: for a significant proportion of patients receiving split care, no communication whatsoever has taken place between their providers of treatment. Indeed, either through directly asking if communication between their two providers had occurred or using the obtaining of consent for communication as a proxy, our results actually yield higher rates of provider non-communication than the previous two surveys (Avena and Kalman, 2010; Kalman et al., 2012). Requesting permission from a patient to speak to the other professional in a split treatment arrangement may reflect a clinician’s intention to communicate with the other professional or, at least, recognition of the possible, future necessity of such contact. Certainly, this request indicates to the patient that the clinicians intend to collaborate about the patient’s care or that such a practice may be a component of care.
Does such communication matter? Does it impact treatment outcome or patient satisfaction? The former issue has never been assessed, just as differential outcomes for split versus integrated forms of combined treatment remain unexplored, and may be logistically impractical to investigate. This study suggests that patient satisfaction with treatment, recognizably important, may indeed be related to provider communication in split-care arrangements.
Response scores to two Likert-scale items, “Overall, how satisfied are you with the split-care arrangements for your mental health treatment?” and “I feel that my Therapist and Prescriber work together in synch to help me” revealed differences between those whose providers had communicated with each other and those for whom no communication had occurred, with the former indicating greater satisfaction. Responses to the first of these items (posed to only the second batch of responders) suggested a statistically insignificant trend toward greater satisfaction among those respondents whose providers had communicated. Responses to the second item revealed significant differences in the degree to which patients feel that their treatment providers worked in synch if communication had taken place. Though intuitively expectable, this data supports continued attention to the importance of communication and the relationship between providers in split care. The second of these two items (“Therapist and Prescriber in Synch”) was intended to investigate a more subjective sense of confidence with treatment than the more direct inquiry about satisfaction. As in other multiple-provider health care situations (e.g., obstetrician/psychopharmacologist; oncologist/surgeon; neurologist/neurosurgeon, etc.), it is generally comforting to patients to know that their respective providers are coordinating their treatments. Beyond patient comfort, certainly one of the significant and oft-touted benefits of Electronic Medical Records (EMRs) is that various providers can readily know what the other is doing and thinking about their mutual patient—allowing for the elimination of redundancies and medical errors, while also providing patients with a sense of everyone being “on board” in their care. In the outpatient mental health world (multiple, unconnected individual providers), EMRs do not exist and direct communication between providers remains the only recourse.
Similarly, our survey results suggest that with respect to a familiarity between prescriber and psychotherapist and the existence of a professional association between them, patients’ comfort with the split-care arrangement is significantly increased for those whose two providers have had communication compared with those whose providers had never spoken.
Outpatients are receiving split care at high rates and the prevalence of this structure is likely increasing. Though both psychiatrists’ and psychologists’ practice guidelines encourage coordination of care, collaborating practitioners are in large part not collaborating. Reasons for non-communication have been extensively considered (Beitman et al., 2003) and are multi-dimensional, including logistical (answering machine phone tag), financial (uncompensated time), psychological (countertransference issues between professionals), and irrational (inaccurate understanding of confidentiality and HIPAA constraints) factors. The current study confirms this practice gap and suggests that patients experience lower rates of satisfaction when they feel their clinicians are not communicating. It is likely that lower patient satisfaction is associated with poorer outcomes, as has been demonstrated previously (Deen et al., 2011; Demyttenaere et al., 2011; Gebhardt et al., 2012; Lee et al., 2008). Patient compliance, critical to treatment outcome, is likely to be enhanced by greater satisfaction with treatment. When done well (i.e., coordinated), split treatment may lead to greater treatment compliance (Balon, 2001), offering the potential for better treatment outcome.
Characteristics of our sample deserve mention. Our respondents included slightly more females than males, averaging 30 years of age, predominantly white and well educated. Psychologists provided a large majority of the psychotherapy, and roughly half of the prescribing professionals were identified as non-psychiatrists. Regarding the representativeness of this sample, though limited data exists on the national split-care population, demographic surveys have revealed that MTurk workers are typically very similar to the national populations from which they are drawn; however, they tend to be younger and have lower income than the national average (Paolacci et al., 2010). Though our population (split-care patients participating in online research via Mechanical Turk) may or may not resemble the split-care population at large, comparisons within our group of respondents remain internally feasible and valid. The use of Mechanical Turk methodology for research in the social sciences, though relatively new, is now widespread. Several studies have also established the methodological viability of this subject pool, demonstrating the reliability and validity of clinical research conducted with an MTurk sample (Shapiro et al., 2013), with comparable internal and external validity to laboratory studies (Horton et al., 2011). Other possible limitations include limited checks on respondent validity or truthfulness and our use of an original, face-valid but nonstandardized questionnaire. In particular, the wording of certain items, such as 18 and 19 (see Appendix), may have biased participant’s responses because of the very nature of the questions; however, our choice of phrasing seems ultimately the most practical for eliciting the information needed in such an exploratory study.
Significant questions remain unexplored in this area and suggest avenues of further inquiry: Are there feasible ways of enhancing provider communication, especially in light of its relationship with patient satisfaction (and possibly treatment compliance and outcome)? Are there identifiable patients for whom communication between providers is not so crucial to their treatment (factors might include diagnosis, personality structure, attachment history)? Conversely, and more critically, are there patients for whom non-communication could be predicted to be problematic and who should not receive split treatment (for example, many borderline patients with patterns of splitting and chaotic interpersonal relationships)?
At a time when a different type of “Integrated Care” (the combining of psychiatric and medical care in new models of health care delivery) is extensively discussed at professional meetings, health care policy forums, and throughout the media, our data on communication in split-care arrangements proves highly relevant. The current study suggests that reducing the gap in provider communication (an issue re-confirmed in this study) may be essential to maximizing the efficacy of such models of care. Newly proposed models of care delivery involve sharing of information and collaboration between providers, often across profoundly different medical subspecialties. If communication between closely allied mental health professionals (psychotherapists and psychopharmacologists) is so difficult to achieve, what are the prospects for communication between professionals in unrelated disciplines whose only links are the patients themselves?
None of the authors have any financial disclosures or conflicts of interest to report.
Avena J, Kalman T (2010) Do psychotherapists speak to psychopharmacologists? A survey of practicing clinicians. J Am Acad Psychoanal Dyn Psychiatry
. 38: 675–683.
Balon R (2001) Positive and negative aspects of split treatment. Psychiatr Ann
. 31: 598–603.
Beitman B, Blinder BJ, Thase ME (2003) Integrating psychotherapy and pharmacotherapy: dissolving the mind-brain barrier
. New York, NY: WW Norton.
Cuijpers P, Sijbrandij M, Koole SL, Andersson G, Beekman AT, Reynolds CF (2014) Adding psychotherapy to antidepressant medication in depression and anxiety disorders: a meta analysis. World Psychiatry
. 13: 56–67.
Deen TL, Fortney JC, Pyne JM (2011) Relationship Between Satisfaction, Patient-Centered Care, Adherence and Outcomes Among Patients in A Collaborative Care Trial for Depression. Adm Policy Ment Health
. 38: 345–355.
Demyttenaere K, Reines EH, Lönn SL, Lader M (2011) Satisfaction with medication is correlated with outcome but not persistence in patients treated with placebo, escitalopram, or serotonin-norepinephrine reuptake inhibitors: a post hoc analysis. Prim Care Companion CNS Disord
. 13. doi:10.4088/PCC.10m01080.
Gebhardt S, Wolak AM, Huber MT (2012) Patient satisfaction and clinical parameters in psychiatric inpatients‚ the prevailing role of symptom severity and pharmacologic disturbances. Compr Psychiatry
. 54: 53–60.
Goin MK (2001) Practical Psychotherapy: Split Treatment: The Psychotherapy Role of the Prescribing Psychiatrist. Psychiatr Serv
. 52: 605–609.
Gutheil TG, Simon RI (2003) Abandonment of patients in split treatment. Harv Rev Psychiatry
. 11: 175–179.
Hansen-Grant S, Riba M (1995) Contact between psychotherapists and psychiatric residents who provide medication backup. Psychiatr Serv
. 46: 774–777.
Horton JJ, Rand DG, Zeckhauser RJ (2011) The online laboratory: Conducting experiments in a real labor market. Exp Econ
. 14: 399–425.
Kalman TP, Kalman VN, Granet R (2012) Do Psychopharmacologists Speak to Psychotherapists? A Survey of Practicing Clinicians. Psychodyn Psychiatry
. 40: 275–285.
Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, Gelenberg AJ, Markowitz JC, Nemeroff CB, Russel JM, Thase ME, Trivedi MH, Blalock JA, Borian FE, Jody DN, DeBattista C, Koran LM, Schatzberg AF, Fawcett J, Hirschfeld RMA, Keitner G, Miller I, Kocsis JH, Kornstein SG, Manber R, Ninan PT, Rothbaum B, Rush AJ, Vivian D, Zajecka J (2000) A comparison of nefazadone, the cognitive behavioral-analysis system of psychotherapy, and their combination for the treatment of chronic depression. N Engl J Med
. 342: 1462–1470.
Lee DS, Tu JV, Chong A, Alter DA (2008) Patient satisfaction and its relationship with quality and outcomes of care after acute myocardial infarction. Circulation
. 118: 1938–1945.
Marcus SC, Olfson M (2010) National trends in the treatment for depression from 1998 to 2007. Arch Gen Psychiatry
. 67: 1265–1273.
Mason W, Suri S (2012) Conducting behavioral research on Amazon’s Mechanical Turk. Behav Res Meth
. 44: 1–23.
Paolacci G, Chandler J, Ipeirotis PG (2010) Running experiments on amazon mechanical turk. Judg Decis Making
. 5: 411–419.
Riba M, Balon R (2008) Combining Psychotherapy and Pharmacotherapy. In Hales R, Yudofsky S, Gabbard GO, eds. The American Psychiatric Publishing Textbook of Clinical Psychiatry
(5th ed). Washington, DC: American Psychiatric Publishing.
Shapiro DN, Chandler J, Mueller PA (2013) Using Mechanical Turk to study clinical populations. Clin Psychol Sci
. 1: 213–220.
United States Department of Health and Human Services (2009) The NSDUH report: Major depressive episode and treatment among adults. In Office of Applied Studies (Ed), Substance Abuse and Mental Health Services
. Rockville, MD.
VandenBos GR, Williams S (2000) Is psychologists’ involvement in the prescribing of psychotropic medication really a new activity? Prof Psychol Res Pract
. 31: 615–618.
Experiences of Patients in Split Treatment
1. Please enter your Mechanical Turk worker ID below. Please enter your ID carefully as this will allow us to approve your work on mturk.
2. What is your age?
3. What is your gender?
c. Other (please specify)
4. What ethnicity do you consider yourself?
a. American Indian or Alaskan Native
b. Hawaiian or other Pacific Islander
c. Non-Hispanic white
f. Asian or Asian American
5. What is your highest educational degree?
a. Did not complete High School
b. High School
c. Some college
d. Completed college
e. Some graduate school
f. Completed graduate school
g. Other (please specify)
6. What is your current employment status?
a. Full-time employment
b. Part-time employment
e. Full-time student
f. Part-time student
g. Other (please specify)
7. My psychotherapist is a
b. Social Worker
d. Pastoral Counselor
e. Other (please specify)
8. My medication prescriber is a
b. Primary Care Physician
c. Other (please specify)
9. Which treatment did you start first?
c. Started both around the same time
10. Before my treatment began, the two professionals treating me knew each other (had either met or spoken personally):
c. Don’t know
11. How did your answer to #10 affect your comfort with the split-care arrangement?
c. Had no effect
12. The two professionals treating me are in the…
a. Same practice group or office suite
b. Same institution
c. Insurance network
d. None of the above
13. How did your answer to #12 affect your comfort level with the split-care arrangement?
c. Had no effect
14. My psychotherapist has asked me for permission (a release) to speak to my prescribing physician
c. Don’t recall
15. My prescribing physician had asked me for permission (a release) to speak to my psychotherapist
c. Don’t recall
16. In the past 6 months my 2 providers (psychotherapist and prescriber) have spoken with each other:
b. One or 2 times
c. More than 2 times
d. Don’t know
17. How did your answer to #16 affect your satisfaction with the split-care arrangement?
c. Had no effect
18. I feel that my Therapist and Prescriber work together in synch to help me
a. Strongly Disagree
c. Neither Agree or Disagree
e. Strongly Agree
19. How did your answer to #18 affect your satisfaction with the split-care arrangement?
c. Had no effect
20. Some psychiatrists also provide psychotherapy as well as prescribe medication
21. Which of the following were the issues for which you sought mental health treatment (check all that apply)
a. Anxiety (panic, phobia)
b. Mood Disorder (depression, bipolar, bereavement)
c. Alcohol/Substance Abuse problems
d. Eating Disorder
e. Personality Disorder
f. Relationship problems (marital/family)
h. Other (please specify)
22. Overall, how satisfied are you with the split-care arrangements for your mental health treatment?
a. Very dissatisfied
e. Very satisfied
23. Are you seeing a psychiatrist/other medical professional for psychiatric medication and a psychotherapist (i.e., psychologist, social worker, counselor, etc.) for psychotherapy?