Skip to main content
NCBI home page
HHS Author Manuscripts logoLink to HHS Author Manuscripts
. Author manuscript; available in PMC: 2025 Jul 27.
Published in final edited form as: Am Psychol. 2024 Apr 18;80(5):729–743. doi: 10.1037/amp0001352

Demographics and Clinical Characteristics of Patients of Prescribing Psychologists, Psychiatrists, and Primary Care Physicians

Phillip M Hughes 1,2,3, Joshua D Niznik 1,4,5, Robert E McGrath 6, Casey R Tak 7, Robert B Christian 8,9, Betsy L Sleath 1,3, Kathleen C Thomas 1,3
PMCID: PMC11881088  NIHMSID: NIHMS2055340  PMID: 38635216

Abstract

Objective:

To describe the characteristics of patients receiving psychotropic medication from prescribing psychologists, psychiatrists, and primary care physicians.

Methods:

This descriptive study was conducted using private insurance claims of patients from New Mexico and Louisiana receiving psychotropic medications (anticonvulsants, antidepressants, antipsychotics, hypotensive agents, anxiolytics/sedatives/hypnotics, and stimulants) from 2004–2021 (n = 307,478). Patient characteristics were captured during the 6 months prior to their first psychotropic medication using administrative information, diagnosis and procedure codes, and medication data. Logistic regression models estimated the associations of patient characteristics with prescriber type. Additional logistic regression models estimated the association of prescriber type with medication classes prescribed.

Results:

Patients were most likely to see specialists (psychologists or psychiatrists) if they had bipolar disorder (average marginal effect (AME) and 95% CI: 0.214 [0.196, 0.231]), schizophrenia/psychotic disorders (0.118 [0.097, 0.138]), or had 1–4 visits of psychotherapy (0.267 [0.258, 0.026]). Specialist patients were most likely to see a prescribing psychologist if they had 1–4 visits of psychotherapy (0.196 [0.183, 0.210]) or had insomnia (0.309 [0.203, 0.415]). Prescribing psychologists were more likely to prescribe antidepressants (0.028 [0.011, 0.045]) and less likely to prescribe antipsychotics (−0.016 [−0.020, −0.012]) than psychiatrists. Primary care physicians were less likely to prescribe all psychotropic medications except antidepressants (0.011 [0.002, 0.019]) and anxiolytics (0.074 [0.067, 0.080]).

Conclusion:

Prescribing psychologists treat patients that are more similar to those of psychiatrists than patients of primary care physicians; they are less likely to prescribe antipsychotics and more likely to prescribe antidepressants.

Keywords: Prescriptive authority, RxP, psychiatrists, primary care physicians, prescribing psychologists

The United States (US) is in the midst of an ongoing mental health crisis, with nearly 1 in 4 adults (22.8%) having a mental illness in 2021 (Substance Abuse and Mental Health Services Administration, 2023). Of those, less than half (47.2%) reported receiving mental health services in the past year (Substance Abuse and Mental Health Services Administration, 2023). A large contributor to this lack of treatment is the long-standing shortage of mental health service providers, particularly among those who can prescribe psychotropic medications (Andrilla et al., 2020; Butryn et al., 2017; Thomas et al., 2009).

As physicians with an expertise in mental health, psychiatrists are typically considered the most highly-trained prescribers in the mental health workforce (Robiner et al., 2020); however, these specialist providers are in extremely limited supply, with over half of US counties lacking a psychiatrist and a provider rate of 15.6 per 100,000 people (Andrilla et al., 2018). Psychiatric nurse practitioners, while also well-trained in mental health, are even scarcer than psychiatrists, with no such providers in two-thirds of US counties and a rate of only 2.1 per 100,000 people (Andrilla et al., 2018). Meanwhile, the need for mental health prescribers has been estimated to be between 16.9 and 25.9 per 100,000 (Andrilla et al., 2018; Konrad et al., 2009). As a result of this prescriber shortage, the vast majority of the psychotropic prescribing done in the US is performed by primary care physicians (Hughes, Annis, et al., 2023; Mark et al., 2009) who receive significantly less training in mental health than specialist providers (Robiner et al., 2020).

Prescriptive authority for psychologists (RxP) has been championed by psychologists since the mid-1980s as a strategy to increase the number of specialist mental health prescribers, gaining momentum following a successful Department of Defense demonstration project in which psychologists were trained to safely prescribe psychotropic medications (DeLeon et al., 1991; Fox, 1988; Fox et al., 2009; Sammons & Brown, 1997). To date, six states (New Mexico, Louisiana, Illinois, Iowa, Idaho, and Colorado) and Guam have passed RxP laws (Licensed Psychologist Prescriptive Authority, 2023; Curtis et al., 2022). Additionally, similar policies enable psychologists to prescribe in the United States Public Health Service, the Indian Health Service, and the Department of Defense (Curtis et al., 2022). The requirements for psychologists seeking an RxP license vary by state regarding practicum requirements, but all prescribing psychologists generally must have a 2-year post-doctoral master’s degree in clinical psychopharmacology and pass a national licensing exam. These requirements align with the educational recommendations laid out by the American Psychological Association Ad Hoc Task Force on Psychopharmacology that was the starting point for the RxP movement within the association (Smyer et al., 1993). As a result of the time involved in these additional requirements, prescribing psychologists remain primarily concentrated in the first two states to pass RxP laws, New Mexico (NM) in 2002 and Louisiana (LA) in 2004 (D. Phillips, personal communication, February 26, 2023). However, the density of prescribing psychologists in those two states had reached 2.5 per 100,000 people by 2019 and has continued growing since then (D. Phillips, personal communication, February 26, 2023; Robiner et al., 2020).

Research on RxP outcomes has been extremely limited. A pair of studies (Linda & McGrath, 2017; Peck et al., 2021) surveyed prescribing psychologists about their practice, finding that the psychologists reported their patients were primarily adults with a mix of insurance coverage. Their participants also reported that the proportion of their case-mix with severe diagnoses, minority status, and rurality all increased after starting prescribing. More recently, a small number of population-based studies have been published that demonstrated reductions in suicides and deaths attributed to mental illness following the adoption of RxP (Choudhury & Plemmons, 2021, 2023; Hughes, McGrath, et al., 2023). An additional study used these estimates to show that RxP policies are a cost-effective measure to reduce suicide rates (Hughes, Phillips, et al., 2023). However, myriad fundamental questions remain. To date, no study has provided objective details of the patient population treated by prescribing psychologists nor the types of medications they are prescribing. Furthermore, no study has compared prescribing psychologists’ patients and prescribing to those of more traditional prescribers – psychiatrists and primary care physicians.

This Study

This study aimed to describe 1) patient demographics and clinical characteristics and 2) the medications prescribed for patients receiving psychotropic medication from prescribing psychologists, psychiatrists, and primary care physicians. We used the Andersen Behavioral Model of Healthcare Utilization (ABM) as the theoretical framework for this study (Andersen, 1995; Babitsch et al., 2012). In this model, an individual’s predisposing factors (e.g., age), enabling factors (e.g., employment status), need factors (e.g., mental health diagnosis), and context factors (e.g., rurality) contribute to the mental health services the individual receives. Using this conceptual model, we estimated the association between patient characteristics and the probability that they receive care from a given type of prescriber. We then estimated the association between prescriber type and the probability of receiving a given psychotropic medication class.

Methods

Study Design, Data, and Sample

We conducted a pooled cross-sectional study of patients with a prescription for psychotropic medication from a psychologist, psychiatrist, or primary care physician. To accomplish this, we used data from the Merative MarketScan Commercial Claims and Encounters Database (henceforth, MarketScan) from 2004–2021. This dataset contains deidentified insurance claims for over 255 million individuals with employer-sponsored insurance from over 350 insurance carriers and is widely used in health research (Butler et al., 2021). Our sample was restricted to patients whose insurance was based in NM or LA due to the limited number of prescribing psychologists in other RxP states. Among those patients in NM or LA, we identified the first claim for a psychotropic medication of at least 14 days’ supply. The date of the claim served as the index date at which patients were described and prescriber type was identified. Psychotropic medications were identified using a previously identified list of medications (Houghton et al., 2017). We used a 14-day requirement to exclude individuals who received only a short-term or one-time psychotropic prescription (e.g., an anxiolytic prior to outpatient surgery); those who did not have a psychotropic prescription for at least 14 days were excluded. We then excluded all individuals who did not have at least 6 months (180 days) of pre-index enrollment and those who did not have continuous enrollment, mental health coverage, and prescription coverage during a 6-month lookback. This allowed for sufficient lookback time for describing patient characteristics and to increase confidence that the index medication was indeed their first recent psychotropic prescription. Finally, patients who did not receive their prescription from a psychologist, psychiatrist, or primary care physician were excluded. A summary of these inclusion and exclusion criteria can be found in Figure 1. All inclusion and exclusion criteria were selected in consultation with a pharmacoepidemiologist, a psychiatrist, and a psychologist. This study was reviewed by the Institutional Review Board at the University of North Carolina at Chapel Hill and determined to be exempt from oversight.

Figure 1. Inclusion and Exclusion Criteria.

Figure 1.

Open in a new tab

Note: Primary care physicians include pediatrics, family medicine, and internal medicine. Patients receiving medication from any other non-psychiatrist physician or an advanced practice provider are excluded from this analysis.

Measures

Prescriber Type

MarketScan uses a provider taxonomy that identifies the provider seen in all outpatient claims. Using this taxonomy, we coded prescribers as psychologists, psychiatrists (including general and child psychiatrists), primary care physicians (including internal medicine, family practice, or pediatricians), or other (all other prescribing taxonomies). However, it is not possible to link prescription claims to the visit at which they were prescribed in the MarketScan data. Given this limitation, it is common practice to identify the prescriber for the most recent outpatient visit prior to the prescription claim as the corresponding prescriber (Farley et al., 2017; McCoul et al., 2019). However, given that a patient may see both a physician (psychiatrist or primary care) and a psychologist when seeking mental health care, we modified the common approach to incorporate a hierarchy of prescribers. Under this approach, we considered all outpatient visits in the 30 days prior to the index date. If a patient ever saw a psychiatrist during that period, we assigned the psychiatrist as the prescriber. If no psychiatrist was seen, we then sequentially considered primary care physicians, then psychologists, and finally all other prescribers. For example, if a patient saw both a primary care physician and a psychologist during the 30-day window, the primary care physician would be assigned as the prescriber.

We conducted a pair of sensitivity analyses relating to identifying prescriber types. First, we used a more conservative approach in our hierarchy in which we considered other prescribers prior to psychologists. Under this specification, only patients who had seen a psychologist and no other prescriber in the month prior to their prescription would be considered as being treated by a psychologist. Second, we used the common approach of assigning the prescriber from the most recent outpatient visit as the prescriber. There was good concordance between our primary specification and both the more conservative approach (73.9%) and the common approach (76.8%). We also examined prescriber identification in 2004, the year before psychologists began prescribing (McGrath, 2010), to explore misclassification in this study and found evidence to suggest that misclassification is not a major concern. Specifically, prescriber identification from 2005–2021 appear to demonstrate divergent validity from the known misclassification of 2004 (see supplementary materials). After this analysis, all individuals with an index date in 2004 were dropped (n = 8,890) to produce a final analytic sample of 307,478 patients.

Psychotropic Medication Type

Psychotropic medications were identified and grouped into six classes using a previously published list of generic psychotropic medication names (Houghton et al., 2017). These medication classes included anticonvulsants, antidepressants, antipsychotics, hypotensive agents, anxiolytics/sedatives/hypnotics, and stimulants.

Andersen Behavioral Model Factors

Patient demographics and clinical characteristics were identified based on their role in the ABM as predisposing, enabling, need, context, or utilization factors. Predisposing factors included age at the index date, age squared, and binary sex as identified on the insurance enrollment file. Enabling factors included employment status (full-time, part-time, retired, other/unknown), employment type (salary, hourly, unknown), and relationship to the employee (employee, spouse, child/dependent).

Need factors included mental health conditions diagnosed during the prior six months (schizophrenia/psychotic disorders, bipolar disorders, depressive disorders, anxiety disorders, post-traumatic stress disorder (PTSD), eating disorders, personality disorders, autism spectrum disorder, attention deficit/hyperactivity disorder (ADHD), conduct disorder), selected physical conditions diagnosed during the prior 6 months (epilepsy/seizure disorders, hypertension, congestive heart failure, liver disease, diabetes, cancer, insomnia), and the Charlson Comorbidity Index (CCI). Specific physical conditions were selected by the psychologist and psychiatrist on the study team due to their potential for increasing the complexity of psychotropic medication management. The CCI is a measure of morbidity that is widely used in claims-based analyses and has been updated to include updated diagnosis codes (Charlson et al., 1994; Glasheen et al., 2019). Given the low variability in CCI scores in our sample, we categorized CCI scores into those with low comorbidity (CCI = 0–2) and elevated comorbidity (CCI = 3 or more). Mental health diagnoses (Stewart et al., 2019) and physical conditions (Amari et al., 2022; Glasheen et al., 2019; Kalilani et al., 2019; Kasman et al., 2020) were both identified using previously published lists of diagnosis codes from the International Classification of Diseases – Clinical Modification (9th and 10th editions as appropriate).

Context factors included rurality as measured by residing in a metropolitan statistical area (MSA; MSA, non-MSA, unknown), state of residence (NM or LA), and the year in which the index date occurred. Finally, healthcare utilization during the six-month lookback period was included in the form of psychotherapy (0 visits, 1–4 visits, 5+ visits) and any psychiatric emergency department (ED) visits. Claims for psychotherapy visits were identified based on the procedure group (individual therapy, family therapy, group therapy, or psychiatric advice) as provided in MarketScan. Psychiatric ED visits were identified as any ED claim for which the primary diagnosis code was a mental health condition as defined above, a standard approach used in national data (Karaca & Moore, 2020).

Analysis

We first examined bivariate descriptive statistics for the ABM factors stratified by prescriber type using chi-squared and Kruskal–Wallis tests, as appropriate. Next, we examined the multivariate association between the ABM factors and prescriber type. The independence of irrelevant alternatives assumption was not met for multinomial logistic regression (Hausman-McFadden test: p<.001); we therefore estimated two logistic regression models: 1) primary care versus mental health specialty care (psychologists and psychiatrists) and 2) psychologists versus psychiatrists. The results of the multinomial logistic regression were largely consistent with the results of this approach (see supplementary materials). Finally, to examine the association between prescriber type and the medication prescribed, we estimated logistic regression models for each of the medication categories with prescriber type and the ABM factors as predictors in the model. This model compared the prescribing of prescribing psychologists and primary care physicians relative to psychiatrists. Given that medication classes are not mutually exclusive in functionality (e.g., bupropion is an antidepressant often prescribed for ADHD), a multinomial logistic regression model intuitively does not meet the independence of irrelevant alternatives assumption. All results are presented as average marginal effects (AMEs) and 95% confidence intervals (CIs).

Transparency and Openness

Data cleaning was conducted in SAS Studio 3.8 (SAS Institute, Cary, NC) and data analysis was performed in Stata 17 (StataCorp, College Station, TX). There was no missingness, allowing for analysis of the full sample. For a complete description of all assumption and specification tests, see the supplementary materials. This study was not pre-registered. The data are proprietary and are therefore not made publicly available; however, all analytic code is available from the corresponding author upon request. We report how we determined our sample size, all data exclusions, all manipulations, and all measures in the study.

Results

Of the 307,478 patients, 4,612 (1.5%) patients received a prescription from a prescribing psychologist, 17,714 (5.8%) from a psychiatrist, and 285,152 (92.7%) from a primary care physician. The sample was predominantly female (59.3%), the insured employee (59.8%), living in an MSA (79.3%), in Louisiana (72.0%), with an average age of 38.2 years (SD = 15.9). The most common mental health conditions included anxiety disorders (10.6%), depressive disorders (8.1%), and ADHD (6.1%), with all others having a prevalence less than 2%. The most common physical conditions included hypertension (16.8%), diabetes (3.4%), and cancer (1.75%), with all others having a prevalence less than 1%. Elevated CCI scores were uncommon (3.7%). All predisposing, enabling, need, context, and utilization factors differed significantly by prescriber type, with the exception of epilepsy (p=0.138; see Table 1).

Table 1.

Patient Predisposing, Enabling, Need, Context, and Healthcare Utilization Factors

Prescribing Psychologists Psychiatrists Primary Care Physicians Total
N (%) N (%) N (%) N (%) p-value
Total N 4,612 17,714 285,152 307,478
Predisposing
Age at Service Date, M (SD) 29.36 (15.63) 31.09 (14.87) 38.80 (15.85) 38.22 (15.94) <.001
Male 2,006 (43.5%) 8,031 (45.34%) 115,151 (40.38%) 125,188 (40.71%) <.001
Enabling
Employment Status <.001
Full-Time 1,174 (25.46%) 5,525 (31.19%) 80,846 (28.35%) 87,545 (28.47%)
Part-Time 28 (0.61%) 88 (0.5%) 1,322 (0.46%) 1,438 (0.47%)
Retired 31 (0.67%) 261 (1.47%) 5,633 (1.98%) 5,925 (1.93%)
Other/Unknown 3,379 (73.27%) 11,840 (66.84%) 19,7351 (69.21%) 21,2570 (69.13%)
Employment Type <.001
Salary 440 (9.54%) 2,123 (11.98%) 27,110 (9.51%) 29,673 (9.65%)
Hourly 415 (9%) 2,083 (11.76%) 35,866 (12.58%) 38,364 (12.48%)
Unknown 3,757 (81.46%) 13,508 (76.26%) 222,176 (77.91%) 239,441 (77.87%)
Relation to Employee <.001
Employee 2,098 (45.49%) 8,825 (49.82%) 173,018 (60.68%) 183,941 (59.82%)
Spouse 636 (13.79%) 2,801 (15.81%) 60,593 (21.25%) 64,030 (20.82%)
Child/Other Dependent 1,878 (40.72%) 6,088 (34.37%) 51,541 (18.07%) 59,507 (19.35%)
Need
Mental Health Diagnoses
Schizophrenia 36 (0.78%) 381 (2.15%) 586 (0.21%) 1,003 (0.33%) <.001
Bipolar Disorders 154 (3.34%) 1,136 (6.41%) 1,006 (0.35%) 2,296 (0.75%) <.001
Depressive Disorders 1,383 (29.99%) 5,053 (28.53%) 18,303 (6.42%) 24,739 (8.05%) <.001
Anxiety Disorders 1,747 (37.88%) 4,650 (26.25%) 26,147 (9.17%) 32,544 (10.58%) <.001
Post-Traumatic Stress Disorder 140 (3.04%) 548 (3.09%) 818 (0.29%) 1,506 (0.49%) <.001
Personality Disorders 277 (6.01%) 662 (3.74%) 2,163 (0.76%) 3,102 (1.01%) <.001
Eating Disorders 27 (0.59%) 117 (0.66%) 306 (0.11%) 450 (0.15%) <.001
Autism Spectrum Disorder 79 (1.71%) 188 (1.06%) 400 (0.14%) 667 (0.22%) <.001
Attention Deficit/Hyperactivity Disorder 1,094 (23.72%) 2,792 (15.76%) 14,741 (5.17%) 18,627 (6.06%) <.001
Conduct Disorder 108 (2.34%) 386 (2.18%) 868 (0.3%) 1,362 (0.44%) <.001
Physical Diagnoses
Epilepsy 8 (0.17%) 64 (0.36%) 944 (0.33%) 1,016 (0.33%) 0.138
Hypertension 197 (4.27%) 997 (5.63%) 5,0467 (17.7%) 51,661 (16.8%) <.001
Diabetes 73 (1.58%) 434 (2.45%) 20,674 (7.25%) 21,181 (6.89%) <.001
Congestive Heart Failure 11 (0.24%) 38 (0.21%) 2,062 (0.72%) 2,111 (0.69%) <.001
Liver Disease 128 (2.78%) 566 (3.20%) 9,769 (3.43%) 10,463 (3.40%) 0.016
Cancer 57 (1.24%) 131 (0.74%) 5,194 (1.82%) 5,382 (1.75%) <.001
Insomnia 48 (1.04%) 32 (0.18%) 528 (0.19%) 608 (0.2%) <.001
Charlson Comorbidity Index 108 (2.34%) 413 (2.33%) 10,972 (3.85%) 11,493 (3.74%) <.001
Context
Rurality <.001
Rural 399 (8.65%) 1,985 (11.21%) 55,625 (19.51%) 58,009 (18.87%)
Metropolitan Statistical Area 4,051 (87.84%) 15,436 (87.14%) 224,197 (78.62%) 243,684 (79.25%)
Unknown 162 (3.51%) 293 (1.65%) 5,330 (1.87%) 5,785 (1.88%)
State of Residence - New Mexico 1,053 (22.83%) 5,597 (31.6%) 79,397 (27.84%) 86,047 (27.98%) <.001
Year <.001
2005 46 (1.00%) 361 (2.04%) 5,953 (2.09%) 6,360 (2.07%)
2006 57 (1.24%) 335 (1.89%) 9,711 (3.41%) 10,103 (3.29%)
2007 47 (1.02%) 242 (1.37%) 8,422 (2.95%) 8,711 (2.83%)
2008 224 (4.86%) 1,127 (6.36%) 17,942 (6.29%) 19,293 (6.27%)
2009 130 (2.82%) 712 (4.02%) 14,149 (4.96%) 14,991 (4.88%)
2010 137 (2.97%) 1,390 (7.85%) 13,653 (4.79%) 15,180 (4.94%)
2011 211 (4.58%) 1,265 (7.14%) 16,286 (5.71%) 17,762 (5.78%)
2012 171 (3.71%) 1,117 (6.31%) 14,236 (4.99%) 15,524 (5.05%)
2013 725 (15.72%) 2,555 (14.42%) 39,134 (13.72%) 42,414 (13.79%)
2014 571 (12.38%) 2,014 (11.37%) 35,127 (12.32%) 37,712 (12.26%)
2015 606 (13.14%) 1,831 (10.34%) 38,724 (13.58%) 41,161 (13.39%)
2016 717 (15.55%) 2,110 (11.91%) 35,000 (12.27%) 37,827 (12.3%)
2017 369 (8.00%) 929 (5.24%) 14,646 (5.14%) 15,944 (5.19%)
2018 150 (3.25%) 432 (2.44%) 6,696 (2.35%) 7,278 (2.37%)
2019 130 (2.82%) 421 (2.38%) 5,728 (2.01%) 6,279 (2.04%)
2020 151 (3.27%) 482 (2.72%) 5,359 (1.88%) 5,992 (1.95%)
2021 170 (3.69%) 391 (2.21%) 4,386 (1.54%) 4,947 (1.61%)
Healthcare Utilization
Psychotherapy <.001
None 1,694 (36.73%) 11,160 (63%) 275,165 (96.5%) 288,019 (93.67%)
1–4 Visits 2,125 (46.08%) 4,701 (26.54%) 6,875 (2.41%) 13,701 (4.46%)
5+ Visits 793 (17.19%) 1,853 (10.46%) 3,112 (1.09%) 5,758 (1.87%)
Any Psychiatric ED visits 31 (0.67%) 262 (1.48%) 380 (0.13%) 673 (0.22%) <.001
Psychotropic Medications <.001
Anticonvulsants 446 (9.67%) 2,220 (12.53%) 41,672 (14.61%) 44,338 (14.42%)
Antidepressants 2,030 (44.02%) 7,671 (43.3%) 122,196 (42.85%) 131,897 (42.9%)
Antipsychotics 132 (2.86%) 1,110 (6.27%) 2,046 (0.72%) 3,288 (1.07%)
Anxiolytics/Sedatives/Hypnotics 536 (11.62%) 2,221 (12.54%) 72,054 (25.27%) 74,811 (24.33%)
Hypotensive Agents 103 (2.23%) 340 (1.92%) 5,349 (1.88%) 5,792 (1.88%)
Stimulants 1,365 (29.6%) 4,152 (23.44%) 41,835 (14.67%) 47,352 (15.4%)
Open in a new tab

Note: Percentages are column percents. Variation in the percentage of patients from each year are partially attributable to insurance plans entering and exiting the MarketScan data and should not be considered a true annual time-trend.

Prescriber Type

The results of our analysis of specialist prescribers compared to primary care physicians can be found in Table 2. Of the predisposing and enabling factors, being male (AME = 0.010; CI [0.009, 0.012]), retired (AME = 0.021; CI [0.012, 0.030]), or a child/dependent (AME = 0.017; CI [0.014, 0.021]) were associated with an increased probability of seeing a specialist, while each additional year of age (AME = −0.0002; CI [−0.0003, −0.0002]) and being an hourly employee (AME = −0.009; CI [−0.012, −0.005]) were associated with an increased probability of seeing a primary care physician. For the need factors, all mental health diagnoses were associated with an increased probability of seeing a specialist, ranging from a 21 percentage-point (pp) higher probability for bipolar disorder (AME = 0.214; CI [0.196, 0.231]) to a 0.7 pp higher probability or personality disorders (AME = 0.007; CI [0.002, 0.013]). Conversely, most physical diagnoses were associated with an increased probability of seeing a primary care physician, including hypertension (AME = −0.035; CI [−0.038, −0.033]), liver disease (AME = −0.007; CI [−0.011, −0.002]), diabetes (AME = −0.022; CI [−0.026, −0.018]), and cancer (AME = −0.011; CI [−0.019, −0.004]). Insomnia was the only physical diagnosis associated with an increased probability of seeing a specialist (AME = 0.023; CI [0.004, 0.042]). Epilepsy, congestive heart failure, and elevated CCI scores were not associated with prescriber type (see Table 2). All context and utilization factors were associated with a higher probability of receiving care from specialists (see Table 2).

Table 2.

Average Marginal Effects and 95% Confidence Intervals for Provider Type

Specialist Care vs Primary Care Prescribing Psychologist vs Psychiatrist
95% CI
 95% CI
AME Lower Upper AME Lower Upper
Predisposing
Age at Service Date* −0.0002 −0.0003 −0.0002 −0.0007 −0.0014 0.0000
Male 0.010 0.009 0.012 −0.016 −0.027 −0.006
Enabling
Employment Status
Part-Time 0.011 −0.002 0.023 0.063 −0.012 0.137
Retired 0.021 0.012 0.030 −0.057 −0.096 −0.017
Other/Unknown −0.001 −0.004 0.002 0.043 0.027 0.059
Employment Type
Hourly −0.009 −0.012 −0.005 0.012 −0.011 0.034
Unknown −0.002 −0.006 0.001 0.014 −0.007 0.036
Relation to Employee
Spouse −0.002 −0.004 0.000 0.002 −0.013 0.018
Child/Other Dependent 0.017 0.014 0.021 0.011 −0.008 0.031
Need
Mental Health Diagnoses
Schizophrenia/Psychotic Disorders 0.118 0.097 0.138 −0.098 −0.131 −0.065
Bipolar Disorders 0.214 0.196 0.231 −0.098 −0.115 −0.080
Depressive Disorders 0.048 0.045 0.052 −0.044 −0.056 −0.032
Anxiety Disorders 0.010 0.008 0.013 0.034 0.022 0.047
Post-Traumatic Stress Disorder 0.043 0.033 0.053 −0.039 −0.066 −0.013
Eating Disorders 0.020 0.005 0.036 −0.065 −0.114 −0.016
Personality Disorders 0.007 0.002 0.013 0.086 0.056 0.115
Autism Spectrum Disorder 0.043 0.028 0.059 −0.028 −0.066 0.010
Attention Deficit/Hyperactivity Disorder 0.039 0.036 0.043 0.033 0.019 0.048
Conduct Disorder 0.024 0.015 0.033 −0.058 −0.086 −0.030
Physical Diagnoses
Epilepsy −0.007 −0.021 0.006 −0.063 −0.149 0.023
Hypertension −0.035 −0.038 −0.033 −0.018 −0.042 0.007
Congestive Heart Failure −0.006 −0.021 0.009 0.061 −0.066 0.188
Liver Disease −0.007 −0.011 −0.002 −0.016 −0.045 0.013
Diabetes −0.022 −0.026 −0.018 −0.039 −0.073 −0.004
Cancer −0.011 −0.019 −0.004 0.095 0.027 0.162
Insomnia 0.023 0.004 0.042 0.309 0.203 0.415
Charlson Comorbidity Index −0.001 −0.007 0.004 0.005 −0.033 0.043
Context
Rurality
Metropolitan Statistical Area 0.024 0.022 0.026 0.023 0.006 0.039
Unknown 0.018 0.012 0.024 0.069 0.030 0.108
State of Residence - New Mexico 0.003 0.001 0.005 −0.055 −0.067 −0.044
Year 0.001 0.001 0.002 0.013 0.011 0.014
Healthcare Utilization
Psychotherapy
1–4 Visits 0.267 0.258 0.275 0.196 0.183 0.210
5+ Visits 0.215 0.204 0.226 0.181 0.161 0.201
Any Psychiatric ED visits 0.040 0.024 0.055 −0.037 −0.089 0.016
Open in a new tab

Note: Specialist care includes prescribing psychologists and psychiatrists. AME = Average Marginal Effect; CI = 95% Confidence Interval; PTSD = post-traumatic stress disorder; ADHD = Attention deficit/hyperactivity disorder; ED = Emergency Department.

*

The average marginal effect of Age includes both Age and Age2.

The results of the specialist versus primary care physician analysis were largely consistent in sensitivity analyses (see supplementary materials). In the conservative hierarchy analysis (psychologists given lowest precedence), personality disorders and insomnia were no longer significant predictors of prescriber type; in the common approach analysis (prescriber assigned on most recent visit before prescription), liver disease and cancer were no longer significant while part-time employment was associated with an increased probability of seeing a specialist (AME = 0.023; CI [0.007, 0.040]).

The results of our analysis of prescribing psychologists compared to psychiatrists can also be found in Table 2. Of the predisposing and need factors, patients were more likely to see a psychiatrist if they were male (AME = −0.016; CI [−0.027, −0.006]) or retired (AME = −0.057; CI [−0.096, −0.017]), and more likely to see a prescribing psychologist if their employment status was other/unknown (AME = 0.043; CI [0.027, 0.059]). Of the need factors, patients were more likely to see a psychiatrist if they had schizophrenia/psychotic disorders (AME = −0.098; CI [−0.131, −0.065]), bipolar disorders (AME = −0.098; CI [−0.115, −0.080]), depressive disorders (AME = −0.044; CI [−0.056, −0.032]), PTSD (AME = −0.039; CI [−0.066, −0.013]), eating disorders (AME = −0.065; CI [−0.114, −0.016]), or conduct disorder (AME = −0.058; CI [−0.086, −0.030]). Patients were more likely to see a prescribing psychologist if they had anxiety disorders (AME = 0.034; CI [0.022, 0.047]), personality disorders (AME = 0.086; CI [0.056, 0.115]), or ADHD (AME = 0.033; CI [0.019, 0.048]). Autism spectrum disorder was not associated with prescriber type. Among the physical conditions, only patients with diabetes were more likely to see a psychiatrist (AME = −0.039; CI [−0.073, −0.004]); conversely, only patients with cancer (AME = 0.095; CI [0.027, 0.162]) or insomnia (AME = 0.309; CI [0.203, 0.415]) were more likely to see prescribing psychologists. No other physical conditions or the CCI were associated with prescriber type. All context and utilization factors were associated with prescribing psychologists with the exceptions of psychiatric ED visits (no association) and residing in New Mexico, which was associated with an increased probability of seeing a psychiatrist (AME = −0.055; CI [−0.067, −0.044]).

The results of the prescribing psychologist versus psychiatrist analysis were largely consistent in the sensitivity analyses (see supplementary materials). In the conservative hierarchy analysis, being male and retired were no longer significant predictors of prescriber type. In addition, epilepsy, hypertension, liver disease, and an elevated CCI were now associated with an increased probability of seeing a psychiatrist (see supplementary materials; note that the changed associations for physical diagnoses can plausibly be attributed to this definition’s requirement that patients of prescribing psychologists saw no other prescriber, making them healthier by definition). In the common approach analysis, being retired and having an eating disorder, diabetes, or cancer were no longer significant while age (AME = −0.0009; CI [−0.0017, −0.00004]) and psychiatric ED visits (AME = −0.077; CI [−0.138, −0.016]) were associated with an increased probability of seeing a psychiatrist.

Psychotropic Medication Type

The results of analyses of medication types can be found in Table 3, including all covariates. Compared to patients who saw a psychiatrist, patients who saw a prescribing psychologist had a higher probability of receiving an antidepressant (AME = 0.028; CI [0.011, 0.045]) and a lower probability of receiving an antipsychotic (AME = −0.016; CI [−0.020, −0.012]). Patients who saw a primary care physician compared to a psychiatrist had lower probabilities of receiving anticonvulsants (AME = −0.025; CI [−0.032, −0.018]), antipsychotics (AME = −0.028; CI [−0.031, −0.026]), hypotensive agents (AME = −0.008; CI [−0.011, −0.005]), and stimulants (AME = −0.014; CI [−0.019, −0.010]), and higher probabilities of receiving antidepressants (AME = 0.011; CI [0.002, 0.019]) and anxiolytics/sedatives/hypnotics (AME = 0.074; CI [0.067, 0.080]). These results were consistent in both sensitivity analyses (see supplementary materials); however, patients of prescribing psychologists were less likely to receive anticonvulsants in both the conservative hierarchy analysis (AME = −0.032; CI [−0.047, −0.016]) and the common approach analysis (AME = −0.018; CI [−0.032, −0.003]).

Table 3.

Average Marginal Effects and 95% Confidence Intervals for Medication Type

Anticonvulsants Antidepressants Antipsychotics
95% CI
 95% CI
 95% CI
AME Lower Upper AME Lower Upper AME Lower Upper
Provider Type
Primary Care Physician v Psychiatrist −0.025 −0.032 −0.018 0.011 0.002 0.019 −0.028 −0.031 −0.026
Prescribing Psychologist v Psychiatrist −0.012 −0.027 0.002 0.028 0.011 0.045 −0.016 −0.020 −0.012
Predisposing
Age at Service Date* 0.0021 0.0020 0.0022 0.0007 0.0005 0.0008 −0.0001 −0.0001 0.0000
Male 0.011 0.008 0.013 −0.077 −0.081 −0.074 0.004 0.003 0.004
Enabling
Employment Status
Part-Time 0.004 −0.014 0.022 −0.016 −0.041 0.008 −0.002 −0.007 0.002
Retired 0.011 0.002 0.019 0.008 −0.005 0.021 0.000 −0.003 0.003
Other/Unknown 0.006 0.002 0.010 −0.018 −0.023 −0.012 0.000 −0.001 0.001
Employment Type
Hourly 0.010 0.004 0.015 0.007 −0.001 0.014 0.003 0.002 0.005
Unknown −0.005 −0.011 0.000 0.008 0.001 0.016 0.002 0.000 0.003
Relation to Employee
Spouse 0.000 −0.003 0.003 −0.007 −0.012 −0.003 0.001 0.000 0.002
Child/Other Dependent −0.024 −0.030 −0.018 0.064 0.056 0.072 0.003 0.002 0.005
Need
Mental Health Diagnoses
Schizophrenia/Psychotic Disorders −0.003 −0.025 0.020 −0.131 −0.159 −0.103 0.092 0.077 0.106
Bipolar Disorders 0.206 0.185 0.227 −0.202 −0.218 −0.185 0.067 0.058 0.075
Depressive Disorders −0.065 −0.070 −0.061 0.237 0.230 0.244 0.004 0.003 0.006
Anxiety Disorders −0.045 −0.049 −0.041 0.047 0.041 0.053 −0.002 −0.003 −0.001
Post-Traumatic Stress Disorder 0.050 0.027 0.074 −0.047 −0.071 −0.022 0.007 0.003 0.011
Eating Disorders −0.030 −0.064 0.004 0.046 0.000 0.092 0.007 −0.001 0.015
Personality Disorders 0.012 −0.005 0.029 −0.015 −0.033 0.004 −0.001 −0.003 0.001
Autism Spectrum Disorder 0.049 0.005 0.093 −0.019 −0.067 0.028 0.050 0.036 0.063
Attention Deficit/Hyperactivity Disorder −0.121 −0.125 −0.118 −0.355 −0.360 −0.350 −0.006 −0.007 −0.006
Conduct Disorder −0.025 −0.055 0.005 −0.003 −0.038 0.031 0.010 0.005 0.014
Physical Diagnoses
Epilepsy 0.648 0.627 0.670 −0.286 −0.309 −0.263 −0.005 −0.008 −0.002
Hypertension 0.020 0.016 0.023 −0.038 −0.043 −0.033 −0.002 −0.003 −0.001
Congestive Heart Failure −0.022 −0.033 −0.010 −0.005 −0.025 0.016 0.000 −0.004 0.005
Liver Disease −0.002 −0.009 0.004 0.002 −0.007 0.011 −0.001 −0.003 0.001
Diabetes 0.064 0.059 0.070 −0.015 −0.022 −0.008 −0.001 −0.002 0.001
Cancer 0.007 −0.002 0.016 −0.054 −0.068 −0.041 0.004 0.000 0.007
Insomnia −0.036 −0.061 −0.012 −0.131 −0.166 −0.096 0.000 −0.008 0.008
Charlson Comorbidity Index 0.026 0.019 0.033 −0.040 −0.049 −0.030 0.005 0.003 0.008
Context
Rurality
Metropolitan Statistical Area −0.004 −0.007 −0.001 0.011 0.007 0.015 −0.001 −0.002 0.000
Unknown −0.023 −0.031 −0.014 0.032 0.019 0.045 0.001 −0.002 0.004
State of Residence - New Mexico 0.007 0.004 0.010 0.052 0.048 0.057 0.001 0.000 0.002
Year 0.007 0.006 0.007 0.001 0.000 0.002 0.000 0.000 0.000
Healthcare Utilization
Psychotherapy
1–4 Visits 0.014 0.005 0.023 0.016 0.006 0.027 0.000 −0.001 0.001
5+ Visits 0.000 −0.012 0.013 0.031 0.017 0.046 0.000 −0.002 0.001
Any Psychiatric ED visits −0.044 −0.073 −0.016 0.050 0.010 0.090 0.006 0.002 0.010
Provider Type
Primary Care Physician v Psychiatrist 0.074 0.067 0.080 −0.008 −0.011 −0.005 −0.014 −0.019 −0.010
Prescribing Psychologist v Psychiatrist 0.006 −0.009 0.020 0.000 −0.005 0.006 −0.003 −0.012 0.006
Predisposing
Age at Service Date* 0.0021 0.0020 0.0023 0.0002 0.0001 0.0002 −0.0044 −0.0045 −0.0043
Male 0.013 0.010 0.016 0.007 0.006 0.008 0.041 0.039 0.043
Enabling
Employment Status
Part-Time 0.014 −0.008 0.036 −0.002 −0.009 0.005 0.004 −0.011 0.018
Retired −0.017 −0.027 −0.007 −0.001 −0.004 0.002 −0.001 −0.012 0.011
Other/Unknown 0.002 −0.003 0.007 0.000 −0.002 0.001 0.010 0.007 0.014
Employment Type
Hourly −0.011 −0.017 −0.004 0.006 0.004 0.008 −0.013 −0.017 −0.009
Unknown −0.007 −0.013 0.000 0.003 0.001 0.005 −0.001 −0.006 0.003
Relation to Employee
Spouse −0.001 −0.005 0.003 0.000 −0.002 0.001 0.006 0.003 0.009
Child/Other Dependent −0.117 −0.123 −0.111 0.007 0.004 0.010 0.037 0.033 0.041
Need
Mental Health Diagnoses
Schizophrenia/Psychotic Disorders −0.049 −0.077 −0.021 −0.005 −0.012 0.003 −0.112 −0.126 −0.097
Bipolar Disorders −0.047 −0.066 −0.028 −0.004 −0.010 0.001 −0.096 −0.103 −0.088
Depressive Disorders −0.108 −0.113 −0.102 −0.012 −0.013 −0.010 −0.067 −0.071 −0.063
Anxiety Disorders 0.095 0.089 0.101 −0.009 −0.010 −0.007 −0.076 −0.078 −0.073
Post-Traumatic Stress Disorder −0.020 −0.042 0.003 0.005 −0.006 0.016 −0.002 −0.025 0.021
Eating Disorders −0.094 −0.133 −0.055 0.007 −0.009 0.023 0.046 0.017 0.074
Personality Disorders 0.029 0.009 0.048 −0.008 −0.014 −0.002 −0.008 −0.022 0.006
Autism Spectrum Disorder −0.099 −0.138 −0.060 0.054 0.039 0.069 −0.081 −0.089 −0.072
Attention Deficit/Hyperactivity Disorder −0.218 −0.222 −0.214 −0.002 −0.004 0.000 0.499 0.491 0.508
Conduct Disorder −0.147 −0.172 −0.122 0.029 0.020 0.038 −0.003 −0.016 0.010
Physical Diagnoses
Epilepsy −0.167 −0.183 −0.151 −0.011 −0.015 −0.006 −0.120 −0.129 −0.111
Hypertension 0.006 0.002 0.010 0.040 0.037 0.042 −0.042 −0.046 −0.039
Congestive Heart Failure 0.002 −0.015 0.019 0.014 0.009 0.019 −0.052 −0.074 −0.030
Liver Disease 0.023 0.015 0.031 −0.006 −0.008 −0.004 −0.017 −0.023 −0.011
Diabetes −0.052 −0.057 −0.046 0.004 0.002 0.006 −0.035 −0.041 −0.029
Cancer 0.047 0.035 0.059 −0.007 −0.010 −0.005 −0.032 −0.045 −0.019
Insomnia 0.225 0.187 0.264 −0.006 −0.015 0.002 −0.056 −0.077 −0.035
Charlson Comorbidity Index 0.011 0.002 0.019 0.004 0.002 0.007 −0.026 −0.033 −0.019
Context
Rurality
Metropolitan Statistical Area −0.008 −0.012 −0.004 −0.003 −0.004 −0.001 0.006 0.003 0.008
Unknown −0.012 −0.024 0.001 −0.004 −0.007 0.000 0.005 −0.002 0.013
State of Residence - New Mexico 0.009 0.006 0.013 −0.005 −0.006 −0.004 −0.076 −0.078 −0.074
Year −0.008 −0.009 −0.008 0.000 −0.001 0.000 0.002 0.001 0.002
Healthcare Utilization
Psychotherapy
1–4 Visits −0.033 −0.042 −0.023 0.005 0.001 0.008 0.007 0.001 0.013
5+ Visits −0.009 −0.023 0.005 0.011 0.005 0.016 −0.013 −0.022 −0.005
Any Psychiatric ED visits 0.037 −0.010 0.084 −0.007 −0.019 0.005 −0.107 −0.124 −0.089
Open in a new tab

Note: AME = Average Marginal Effect; CI = 95% Confidence Interval; PTSD = post-traumatic stress disorder; ADHD = Attention deficit/hyperactivity disorder; ED = Emergency Department.

*

The average marginal effect of Age includes both Age and Age2.

Discussion

Prescribing psychologists are a small but growing workforce, and research to date, although limited, suggests that they positively impact population mental health (Choudhury & Plemmons, 2023; Hughes, McGrath, et al., 2023; Hughes, Phillips, et al., 2023). The present study represents the next step in generating a robust body of evidence regarding prescribing psychology by providing the first individual-level description of the patients seen by prescribing psychologists and the medications they receive. Furthermore, our modeling results provide a first step towards clarifying the role of prescribing psychologists in the mental health workforce in terms of how their patient case-mix and initial prescribing choices compare to those of psychiatrists and primary care physicians. Given the relatively limited literature that exists on prescribing psychologists (Choudhury & Plemmons, 2023; Hughes, McGrath, et al., 2023; Hughes, Phillips, et al., 2023; Linda & McGrath, 2017; Peck et al., 2021), these findings suggest a myriad of future directions and potential policy implications.

The results of our models and the assumption test demonstrating that prescribing psychologists and psychiatrists are statistically similar alternatives (i.e., failed independence of irrelevant alternatives test) provide evidence that prescribing psychologists are indeed treating a patient population similar to that of psychiatrists. They appear to be functioning as a direct complement to psychiatrists to provide specialized mental health care as opposed to filling a more niche role between psychiatrists and primary care physicians. While we found that patients with severe mental health conditions (e.g., schizophrenia/psychotic disorders) were less likely to see a prescribing psychologist than a psychiatrists, prescribing psychologists and psychiatrists appear to be treating patients with similar levels of physical comorbidity. This suggests that prescribing psychologists are serving patients who need specialty mental health care but would have otherwise been seen by a primary care physician, potentially increasing access to care and reducing wait times for specialist appointments. Future research is needed to examine the impact of RxP on access and appointment wait times for psychiatrists and psychologists. Additionally, research focused on patient experiences could identify where prescribing psychologists’ patients would have otherwise sought services in order to understand how RxP is impacting the overall mental health workforce. Finally, the absence of major differences in physical comorbidities may at least partially reflect the medically-oriented clinical settings in which many prescribing psychologists practice. A survey of prescribing psychologists found that 28% were prescribing in community health clinics, 7% in general hospital settings, and 11.6% in private medical practices (Peck et al., 2021). Future research should seek to compare the clinical characteristics of prescribing psychologists’ patients across various clinical settings to clarify instances where prescribing psychologists may be serving different populations.

The difference in mental health severity between patients of prescribing psychologists and psychiatrists also warrants further study to examine how these differences arise. A survey of prescribing psychologists found that overall patient severity increased for psychologists when they began prescribing (Linda & McGrath, 2017), suggesting that although on average psychiatrists typically see more severe patients, the gap in patient severity between psychologists and psychiatrists may be more narrow for prescribing psychologists. As such, prescribing psychologists may be relieving some of the strain on psychiatrists by treating more severe patients than they were prior to RxP. This potential shift in workforce strain should be examined in future research on workforce outcomes such as burnout and job satisfaction. From a health economics lens, future research is also needed to understand the rates with which different mental health prescribers engage in financially motivated patient selection practices, such as selectively treating healthier rather than sicker patients (i.e., ‘creaming’) and avoidance of more severe patients (i.e., ‘dumping’) (Ellis, 1998; Jacobs, 2014). Additionally, given that the present study focused on new patients, future research is needed to understand how the pre-existing patient population may have been impacted by RxP adoption. It seems plausible that the inclusion of prevalent service users may amplify the differences we identified between prescribers.

The impact of RxP on access to mental health services for underserved patients, such as those in rural areas, has long been debated (Curtis et al., 2022). We found that patients located in an MSA were more likely to see a prescribing psychologist, meaning a smaller proportion of their patients are from rural areas. This is in direct contrast to a survey of prescribing psychologists in which 40% of respondents reported that they were treating more patients from rural areas (Linda & McGrath, 2017). It is plausible that prescribing psychologists see fewer patients from rural areas than urban areas but see more patients from rural areas than they did before becoming prescribers. Additionally, private insurance rates are lower in rural versus urban areas (Gong et al., 2019; Long et al., 2021). The pressures of low private coverage and greater unmet need among publicly insured or uninsured patients may result in prescribing psychologists who practice in rural areas seeing more patients who are not covered by private insurance. Future research using Medicaid and Medicare claims is necessary to bring this issue into focus.

Another major issue often raised regarding RxP has been safety, especially regarding the prescribing of antipsychotics given the heightened risk profile of those particular medications (Curtis et al., 2022; Robiner et al., 2013). While this study was not designed to address questions of safety, our findings provided tangential evidence in this area. Specifically, we found that patients of prescribing psychologists are more likely to receive antidepressants and less likely to receive antipsychotics than patients of psychiatrists, even after accounting for clinical characteristics. This aligns with national psychotropic prescribing trends (Hughes, Annis, et al., 2023), where antipsychotics are the only psychotropic medication for which psychiatrists are the majority prescriber. Future research is needed to understand the source of this difference in prescribing practice. The difference in medication choice may reflect the more multimodal approach to treatment used by psychologists, where psychotherapy is being supplemented by less intensive medications; alternatively, given that these are initial prescriptions and not necessarily the final medication the patient ends up receiving, this could reflect a more conservative approach to medication initiation among the psychologists, in which they start with less intensive medications and adjust as needed. Qualitative research with prescribing psychologists is needed to better elucidate the rationale for this difference in prescribing.

Limitations

This study has limitations worth noting. First, we had to rely on an algorithmic approach to identifying prescribers due to the absence of prescriber information on medication claims in these data. This limitation warrants significant consideration, as prescriber identification was a critical component of the present study; however, our examination of prescribers in 2004 and the consistency of our results across sensitivity analyses using different identification methods instill confidence that our results are robust to misclassification related to this limitation. Second, this study is subject to the standard limitations associated with claims data. We were unable to account for patient race/ethnicity in our analysis, as this information was not available in claims. Future studies should seek to examine the distribution of race/ethnicity between prescribers in this population. Similarly, we were unable to see office visits for which insurance was not billed. We anticipate that this primarily impacted psychotherapy visits, but it is also possible that some individuals paid out-of-pocket for visits where they received a prescription for psychotropic medications. Such a scenario could result in their prescription being misattributed to a different prescriber type.

Finally, while we attempted to account for temporal differences in the proportion of patients treated by different providers using year fixed effects, some temporal issues may be worth additional consideration. It is possible that the timing with which private insurance began to cover prescribing psychologists may result in their patients being underrepresented in the earlier years of this study. However, a prior study found that private insurance is one of the major reimbursement sources for prescribing psychologists (Peck et al., 2021), suggesting that this limitation is not cause for concern. Similarly, there have been substantial changes to private insurance during the study period, such as the passage of the Affordable Care Act and various mental health parity acts, all of which could impact what providers patients may be able to see. A future study is necessary to examine this further using a quasi-experimental design. Furthermore, the availability of training programs may have limited the roll-out of RxP in LA during the first few years of their policy, perhaps resulting in early years of the study period primarily reflecting patients from NM.

Conclusion

This study provides a foundational understanding of care delivered by prescribing psychologists versus other healthcare providers, providing a path forward for patient outcomes research to better understand their role in the mental health workforce. Among privately insured patients, prescribing psychologists appear to be treating a patient population similar to that of psychiatrists, and distinct from patients that would otherwise be treated by primary care physicians. Patients of prescribing psychologists are less likely to receive antipsychotic medications even after adjusting for clinical characteristics, suggesting that significant professional differences exist regarding how decisions for initial medication treatment are made.

Supplementary Material

supplemental material

Public Significance Statement:

This study provides the first objective description of the patients receiving treatment from prescribing psychologists. Prescribing psychologists have similar patients to psychiatrists, suggesting they are serving patients in need of specialty mental health services. Prescribing psychologists are, however, also less likely to prescribe antipsychotics than psychiatrists, even after adjusting for patient clinical characteristics. This may suggest that prescribing psychologists take a more conservative approach to prescribing certain psychotropic medications.

Funding:

This research was partially supported by a National Research Service Award Pre-Doctoral/Post-Doctoral Traineeship from the Agency for Healthcare Research and Quality sponsored by The Cecil G. Sheps Center for Health Services Research, The University of North Carolina at Chapel Hill, Grant No. T32-HS000032. The funder played no role in the planning, conducting, writing, or decision to publish the study.

Footnotes

Conflicts of Interest: Phillip Hughes was awarded the 2023 Patrick H. DeLeon Prize for Outstanding Student Contribution to the Advancement of Pharmacotherapy from APA Division 55 (Society for Prescribing Psychology). No other authors had conflicts to disclose.

Data Availability:

This study was not pre-registered. The data are proprietary and are therefore not made publicly available; however, all analytic code is available from the corresponding author upon request.

References

  1. Amari DT, Juday TR, Frech FH, Wang W, Gor D, Atkins N, & Wickwire EM (2022). Fall Risk, Healthcare Resource Use, and Costs Among Adult Patients in the United States Treated for Insomnia with Zolpidem, Trazodone, or Benzodiazepines: A Retrospective Cohort Study. Advances in Therapy, 39(3), 1324–1340. 10.1007/s12325-022-02041-4 [DOI] [PubMed] [Google Scholar]
  2. Andersen RM (1995). Revisiting the Behavioral Model and Access to Medical Care: Does it Matter? Journal of Health and Social Behavior, 36(1), 1. 10.2307/2137284 [DOI] [PubMed] [Google Scholar]
  3. Andrilla CHA, Garberson LA, Patterson DG, Quigley TF, & Larson EH (2020). Comparing the Health Workforce Provider Mix and the Distance Travelled for Mental Health Services by Rural and Urban Medicare Beneficiaries. The Journal of Rural Health : Official Journal of the American Rural Health Association and the National Rural Health Care Association. 10.1111/jrh.12504 [DOI] [PubMed] [Google Scholar]
  4. Andrilla CHA, Patterson DG, Garberson LA, Coulthard C, & Larson EH (2018). Geographic Variation in the Supply of Selected Behavioral Health Providers. American Journal of Preventive Medicine, 54(6), S199–S207. 10.1016/j.amepre.2018.01.004 [DOI] [PubMed] [Google Scholar]
  5. Babitsch B, Gohl D, & von Lengerke T (2012). Re-revisiting Andersen’s Behavioral Model of Health Services Use: A systematic review of studies from 1998–2011. Psycho-Social Medicine, 9, Doc11. 10.3205/psm000089 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Butler AM, Nickel KB, Overman RA, & Brookhart MA (2021). IBM MarketScan Research Databases. In Databases for Pharmacoepidemiological Research (pp. 243–251). Springer. [Google Scholar]
  7. Butryn T, Bryant L, Marchionni C, & Sholevar F (2017). The shortage of psychiatrists and other mental health providers: Causes, current state, and potential solutions. International Journal of Academic Medicine, 3(1), 5. Gale OneFile: Health and Medicine. 10.4103/IJAM.IJAM_49_17 [DOI] [Google Scholar]
  8. Charlson M, Szatrowski TP, Peterson J, & Gold J (1994). Validation of a combined comorbidity index. Journal of Clinical Epidemiology, 47(11), 1245–1251. 10.1016/0895-4356(94)90129-5 [DOI] [PubMed] [Google Scholar]
  9. Choudhury AR, & Plemmons A (2021). Deaths of Despair: Prescriptive Authority of Psychologists and Suicides. The Center for Growth and Opportunity at Utah State University. https://www.thecgo.org/research/deaths-of-despair/ [Google Scholar]
  10. Choudhury AR, & Plemmons A (2023). Effects of giving psychologists prescriptive authority: Evidence from a natural experiment in the United States. Health Policy, 134, 104–846. 10.1016/j.healthpol.2023.104846 [DOI] [PubMed] [Google Scholar]
  11. Licensed Psychologist Prescriptive Authority, HB23–1071, 2023 Regular Session (2023). http://leg.colorado.gov/bills/hb23-1071 [Google Scholar]
  12. Curtis SE, Hoffmann S, & O’Leary Sloan M (2022). Prescriptive authority for psychologists: The next step. Psychological Services. 10.1037/ser0000667 [DOI] [PubMed] [Google Scholar]
  13. DeLeon PH, Fox RE, & Graham SR (1991). Prescription privileges: Psychology’s next frontier? American Psychologist, 46(4), 384–393. 10.1037/0003-066X.46.4.384 [DOI] [PubMed] [Google Scholar]
  14. Ellis RP (1998). Creaming, skimping and dumping: Provider competition on the intensive and extensive margins. Journal of Health Economics, 17(5), 537–555. 10.1016/S0167-6296(97)00042-8 [DOI] [PubMed] [Google Scholar]
  15. Farley JF, Hansen RA, Domino ME, Borse M, Mahendraratnam N, Ray N, & Maciejewski ML (2017). Continuity of medication management in Medicaid patients with chronic comorbid conditions: An examination by mental health status. General Hospital Psychiatry, 45, 25–31. 10.1016/j.genhosppsych.2016.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fox RE (1988). Prescription privileges: Their implications for the practice of psychology. Psychotherapy: Theory, Research, Practice, Training, 25(4), 501. [Google Scholar]
  17. Fox RE, DeLeon PH, Newman R, Sammons MT, Dunivin DL, & Baker DC (2009). Prescriptive authority and psychology: A status report. American Psychologist, 64(4), 257–268. 10.1037/a0015938 [DOI] [PubMed] [Google Scholar]
  18. Glasheen WP, Cordier T, Gumpina R, Haugh G, Davis J, & Renda A (2019). Charlson Comorbidity Index: ICD-9 Update and ICD-10 Translation. American Health & Drug Benefits, 12(4), 188–197. [PMC free article] [PubMed] [Google Scholar]
  19. Gong G, Phillips SG, Hudson C, Curti D, & Philips BU (2019). Higher US Rural Mortality Rates Linked To Socioeconomic Status, Physician Shortages, And Lack Of Health Insurance. Health Affairs, 38(12), 2003–2010. 10.1377/hlthaff.2019.00722 [DOI] [PubMed] [Google Scholar]
  20. Houghton R, Ong RC, & Bolognani F (2017). Psychiatric comorbidities and use of psychotropic medications in people with autism spectrum disorder in the United States: Psychotropic medications in autism in the US. Autism Research, 10(12), 2037–2047. 10.1002/aur.1848 [DOI] [PubMed] [Google Scholar]
  21. Hughes PM, Annis IE, McGrath RE, & Thomas KC (2023). Psychotropic Medication Prescribing Across Medical Providers, 2016–2019. Psychiatric Services, appi.ps.20230156. 10.1176/appi.ps.20230156 [DOI] [PubMed] [Google Scholar]
  22. Hughes PM, McGrath RE, & Thomas KC (2023). Evaluating the impact of prescriptive authority for psychologists on the rate of deaths attributed to mental illness. Research in Social and Administrative Pharmacy, 19(4), 667–672. 10.1016/j.sapharm.2022.12.006 [DOI] [PubMed] [Google Scholar]
  23. Hughes PM, Phillips DC, McGrath RE, & Thomas KC (2023). Examining Psychologist Prescriptive Authority as a Cost-Effective Strategy for Reducing Suicide Rates. Professional Psychology: Research and Practice, 54(4), 284–294. 10.1037/pro0000519 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jacobs R (2014). Payment by results for mental health services: Economic considerations of case-mix funding. Advances in Psychiatric Treatment, 20(3), 155–164. 10.1192/apt.bp.113.011312 [DOI] [Google Scholar]
  25. Kalilani L, Friesen D, & Murray P (2019). Treatment patterns in patients with a new diagnosis of epilepsy and psychiatric comorbidities. Epilepsy & Behavior, 99, 106405. 10.1016/j.yebeh.2019.07.006 [DOI] [PubMed] [Google Scholar]
  26. Karaca Z, & Moore BJ (2020). Costs of Emergency Department Visits for Mental and Substance Use Disorders in the United States, 2017: Statistical Brief# 257 (Rev. 2020). In Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. Agency for Healthcare Research and Quality (US). [PubMed] [Google Scholar]
  27. Kasman AM, Zhang CA, Li S, Stevenson DK, Shaw GM, & Eisenberg ML (2020). Association of preconception paternal health on perinatal outcomes: Analysis of U.S. claims data. Fertility and Sterility, 113(5), 947–954. 10.1016/j.fertnstert.2019.12.026 [DOI] [PubMed] [Google Scholar]
  28. Konrad TR, Ellis AR, Thomas KC, Holzer CE, & Morrissey JP (2009). County-Level Estimates of Need for Mental Health Professionals in the United States. Psychiatric Services, 60(10), 1307–1314. 10.1176/ps.2009.60.10.1307 [DOI] [PubMed] [Google Scholar]
  29. Linda WP, & McGrath RE (2017). The current status of prescribing psychologists: Practice patterns and medical professional evaluations. Professional Psychology: Research and Practice, 48(1), 38–45. 10.1037/pro0000118 [DOI] [Google Scholar]
  30. Long JC, Delamater PL, & Holmes GM (2021). Which Definition of Rurality Should I Use?: The Relative Performance of 8 Federal Rural Definitions in Identifying Rural-Urban Disparities. Medical Care, 59(Suppl 5), S413–S419. 10.1097/MLR.0000000000001612 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mark TL, Levit KR, & Buck JA (2009). Datapoints: Psychotropic Drug Prescriptions by Medical Specialty. Psychiatric Services, 60(9), 1167–1167. 10.1176/ps.2009.60.9.1167 [DOI] [PubMed] [Google Scholar]
  32. McCoul ED, Weinreich HM, Mulder H, Man L-X, Schulz K, & Shin JJ (2019). Health Care Utilization and Prescribing Patterns for Adult Eustachian Tube Dysfunction. Otolaryngology–Head and Neck Surgery, 160(6), 1071–1080. 10.1177/0194599819826959 [DOI] [PubMed] [Google Scholar]
  33. McGrath RE (2010). Prescriptive Authority for Psychologists. Annual Review of Clinical Psychology, 6(1), 21–47. 10.1146/annurev-clinpsy-090209-151448 [DOI] [PubMed] [Google Scholar]
  34. Peck KR, McGrath RE, & Holbrook BB (2021). Practices of prescribing psychologists: Replication and extension. Professional Psychology: Research and Practice, 52(3), 195–201. 10.1037/pro0000338 [DOI] [Google Scholar]
  35. Phillips D (2023, February 26). Personal Communication [Personal communication; ]. [Google Scholar]
  36. Robiner WN, Tompkins TL, & Hathaway KM (2020). Prescriptive authority: Psychologists’ abridged training relative to other professions’ training. Clinical Psychology: Science and Practice, 27(1). 10.1111/cpsp.12309 [DOI] [Google Scholar]
  37. Robiner WN, Tumlin TR, & Tompkins TL (2013). Psychologists and medications in the era of interprofessional care: Collaboration is less problematic and costly than prescribing. Clinical Psychology: Science and Practice, 20(4), 489–507. 10.1111/cpsp.12054 [DOI] [Google Scholar]
  38. Sammons MT, & Brown AB (1997). The Department of Defense Psychopharmacology Demonstration Project: An evolving program for postdoctoral education in psychology. Professional Psychology: Research and Practice, 28(2), 107–112. 10.1037/0735-7028.28.2.107 [DOI] [Google Scholar]
  39. Smyer MA, Balster RL, Egli D, Johnson DL, Kilbey MM, Leith NJ, & Puente AE (1993). Summary of the report of the Ad Hoc Task Force on Psychopharmacology of the American Psychological Association. Professional Psychology: Research and Practice, 24(4), 394. [Google Scholar]
  40. Stewart CC, Lu CY, Yoon TK, Coleman KJ, Crawford PM, Lakoma MD, & Simon GE (2019). Impact of ICD-10-CM Transition on Mental Health Diagnoses Recording. eGEMs (Generating Evidence & Methods to Improve Patient Outcomes), 7(1), 14. 10.5334/egems.281 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Substance Abuse and Mental Health Services Administration. (2023). 2021 NSDUH Detailed Tables [National Report]. Substance Abuse and Mental Health Services Administration. https://www.samhsa.gov/data/report/2021-nsduh-detailed-tables [Google Scholar]
  42. Thomas KC, Ellis AR, Konrad TR, Holzer CE, & Morrissey JP (2009). County-Level Estimates of Mental Health Professional Shortage in the United States. Psychiatric Services, 60(10), 1323–1328. 10.1176/ps.2009.60.10.1323 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental material
NIHMS2055340-supplement-supplemental_material.docx (118.3KB, docx)

Data Availability Statement

This study was not pre-registered. The data are proprietary and are therefore not made publicly available; however, all analytic code is available from the corresponding author upon request.

RESOURCES