Abstract
This research highlights the prevalence of burnout and perceived pressures related to productivity requirements among occupational therapy practitioners working in geriatric settings.
Burnout is “a significant occupational hazard that can impair physical health, psychological well-being, and work performance” (Maslach & Leiter, 2008, p. 498). Maslach and Leiter (2022) have conceptualized burnout as including three constructs: extreme emotional exhaustion (EE); extreme depersonalization (DP), and low levels of personal accomplishment (PA). EE is often associated with high workloads, DP results in disengagement and cynicism toward one’s work, and low PA results in feelings of inefficacy. In recent years, practitioner stress may have been complicated by changing payment models and a global pandemic (Maslach & Leiter, 2022). The patient-driven payment model (PDPM), in effect since 2019, projects patient needs on the basis of diagnosis rather than therapist evaluation, removing therapy as the determinant for reimbursement in skilled nursing facilities (SNFs; “Medicare Program,” 2019). Changes in the PDPM led employers to increase revenue through increased productivity standards and decreased staff while maintaining or increasing caseloads (Warren & Sampson, 2020). Neither the PDPM nor the Centers for Medicare and Medicaid Services (CMS) Medicare Benefit Policy Manual acknowledges productivity standards; they are both corporate conventions for revenue purposes. Increased productivity standards have contributed to unethical practices and burnout, which correlate to decreased patient outcomes and practitioner quality of life (Delgadillo et al., 2018).
The purpose of this study was to examine productivity standards and burnout in the context of setting and role, as reported by occupational and physical therapy practitioners as well as speech-language pathologists in geriatric settings, and to explore relationships between productivity standards and perceived ethical pressures. The core constructs of Maslach and Leiter’s studies of burnout (Maslach, 2018; Maslach & Leiter, 2008; Maslach et al., 2012) serve as an organizing framework for this study.
Burnout
Burnout may be experienced when a job requires continuous physical or psychological demands that exceed an individual’s available capacity. Job demands include “work frequency and duration, work intensity, and the nature of the work itself, including the context in which the work is performed” (National Academies of Sciences, Engineering, and Medicine [NASEM], 2019, p. 83). When individuals experience burnout, their job performance becomes affected. Employees make errors, complete the minimum requirements for their position, lack creativity, and decrease standards of performance (Maslach, 2018). For health care professionals, errors may trickle down and affect the quality of patient care. The Work Institute Retention Report (2020) cites a turnover rate increase of 88% since 2010, with the top reasons for leaving listed as job characteristics (10%), poor management (12%) and decreased work life balance (12%). We may see “one out of every three workers choosing to leave their jobs by 2023” (Work Institute Retention Report, 2020, p. 7), which will be costly, because turnover costs are approximately 30% of an employee’s salary, with no return on investment if the employee quits within the first year (Work Institute Retention Report, 2020).
Negative burnout effects include cardiovascular problems, headaches, gastrointestinal disorders, stress, exhaustion, hypertension, depression, and anxiety (Maslach, 2018). Five percent to 8% of health care costs annually correlate to stressors at work (Bostock, 2019), causing increased turnover rates, decreased performance and absenteeism. A meta-analysis of 82 studies correlated burnout levels in health care workers to negative patient outcomes such as decreased satisfaction, perception of safety, and perceived quality of care (Salyers et al., 2017).
Zambo Anderson et al. (2015) examined burnout in 1,366 physical therapists, with 29% who were found to be emotionally exhausted, 15% who had high-perceived stress, and 13% who were experiencing burnout (Zambo Anderson et al., 2015). Workload was found to be a main predictor of EE in occupational therapists, accounting for “29.9% of the variance in the exhaustion trait” of the Maslach Burnout InventoryTM (MBI; Gupta et al., 2012, p. 88). High levels of burnout are aligned with perceived low supervisory support, inadequate income, and limited educational attainment (Shin et al., 2022). In conjunction with burnout, coronavirus disease 2019 (COVID-19) was a recent global event that affected many health care workers. Sharifi et al.’s (2021) systematic review of burnout during COVID-19 identified stressors contributing to burnout as lack of COVID-19 testing and personal protective equipment, rapid changes in health policies, conflicting information, and decreased income.
Productivity and Ethics
Occupational therapists, physical therapists, and speech-language pathologists are required to maintain high caseloads while providing quality care in less time than previously allotted (American Occupational Therapy Association Advisory Opinion for the Ethics Commission [AOTAEC], 2023). Employers often track a practitioner’s job efficiency through a productivity standard, which is a percentage indicating the daily minutes an employee is billing for patients divided by the daily total time on the job. Billable time is actual minutes spent providing skilled services to a patient (CMS, 2019). The higher the productivity standard, the less time a therapist has to address other job demands, such as finishing documentation that was not completed with the patient, chart reviews, discussions with family and nursing staff while the patient is not present, interdisciplinary meetings, cleaning equipment, and other ancillary tasks (CMS, 2019). Despite employer compliance programs, practitioners are responsible to ensure that legal and billing practice information that is provided to them is accurate (AOTAEC, 2023). Data from 3,446 physical therapy practitioners determined that practitioners at SNFs had the highest rate of productivity standards compared with those in other rehabilitation settings (Tammany et al., 2019). Productivity rates positively correlated with the incidence of unethical behaviors (Tammany et al., 2019). Workload expectations are an appropriate way to maximize reimbursement and staffing resources, yet they must be reasonable, ethical, and legal (AOTAEC, 2023). As productivity expectations rise, workload concomitantly increases and autonomy and clinical supervision decrease (Delgadillo et al., 2018).
The American Speech-Language-Hearing Association (ASHA), in surveying more than 4,500 members working in SNFs post-PDPM, documented increased managerial pressures to provide group and concurrent treatment regardless of clinical judgment or patient needs. Escalating administrative mandates from PDPM pressured 60% of respondents to violate the ASHA Code of Ethics (Warren & Sampson, 2020).
Outright fraudulent behavior for financial gain was documented in a January 2016 court case when Kindred/Rehabcare settled for more than $125 million because of False Claims Act allegations (U.S. Department of Justice [DOJ], 2018). Kindred/Rehabcare was accused of setting unrealistic financial goals to maximize reimbursement without considering patient needs (DOJ, 2018). Company tactics used, as per the DOJ report, were keeping patients on caseload longer than necessary, recording minutes practiced by one discipline to another to ensure target reimbursement rates, documenting treatment time in excess of the evaluation code, documenting skilled therapy when a patient was sleeping or unable to benefit, and rounding up treatment minutes (DOJ, 2018). Billing for medically unnecessary services exploits vulnerable adults and decreases the merit of Medicare services (DOJ, 2017). In recognition of, and response to, growing concerns regarding these issues, this study addresses the contextual experience of productivity, burnout, and ethical pressures for practitioners in geriatric settings.
Method
Design
This study used a cross-sectional, mixed-methods survey design. This report includes only the quantitative responses.
Participants
Initial respondents (n = 416), who were recruited through convenience sampling, met the inclusion criteria of being a licensed occupational therapy practitioner, physical therapy practitioner, or speech-language pathologist; being proficient in English; residing in the United States; and working in a SNF or an assisted living facility (ALF). Rehabilitation directors (12.3%) were removed from analysis because of the significant differences of productivity standards compared with those of practitioners. The demographic characteristics of the 366 remaining respondents are presented in Table 1.
Participant Demographics
Note. Racial classifications were self-selected by participants. ALF = assisted living facility; SNF = skilled nursing facility.
Instrument
Part 1 of the survey included the 22 questions of the MBI: Human Services Survey for Medical Personnel (MBI–HSS [MP]), representing three subscales that measure EE, DP, and PA (Maslach, 2018). Burnout is associated with high levels of EE, high levels of DP, and low levels of PA. Participants respond to each question on the MBI–HSS (MP) using a 7-point Likert-type scale ranging from 0 (never) to 6 (always). For this study, cutoff scores to determine extremes in each of the dimensions were set at EE ≥ 27, DP ≥ 10, and PA ≤ 33, consistent with those used by Brady et al. (2020). Current best practices also include creating profile scores (Maslach & Leiter, 2021) on the basis of subscale scores. These profiles include burnout (extreme scores on EE, DP, and PA); overextended (extreme score on EE only), ineffective (extreme score on PA only), disengaged (extreme score on DP only), and engagement (favorable scores on EE, DP, and PA; Maslach & Leiter, 2021).
Observed reliabilities (Cronbach’s α) of the MBI–HSS (MP) and its subscales were all within acceptable ranges (α > .70). The reliabilities of the subscales were as follows: EE, α = .90; DP, α = .76; and PA, α = .72. The reliability of the whole instrument was α = .89. Reliability statistics are not appropriate for any other measures, because no other instruments in this study use agglomerated scoring (i.e., mean or sum scores).
Part 2 of the survey required reporting a productivity standard (either number or range); whether it was reasonable; and if not, what an appropriate standard would be. A productivity standard is classified as the total number of session minutes that a therapist billed for patients divided by the number of minutes that a therapist spends in the building on a single day. Respondents reported the frequency of perceived pressures to enact ethically questionable behaviors, including ▪ changing clock time, ▪ keeping patients on caseload longer than necessary, ▪ increasing session length, ▪ falsifying patient documentation, ▪ changing billing codes, and ▪ billing for patient refusals.
To report frequency, we used a 5-point scale on which the responses were 1 = never, 2 = infrequently, 3 = frequently, 4 = consistently, and 5 = weekly/daily.
Procedure
The study was approved by the Misericordia University Institutional Review Board (No. 2021-0150), and the informed consent of participants was obtained. The survey was shared with various Facebook groups on administrative approval, including the Alternative Healthcare Career for Rehabilitation Professionals, Geriatric OT, PT and SLP Collaborative Group, Misericordia University Occupational Therapy Department, American Therapy Alliance, Therapists for Unionizing, and Occupational Therapy New Grads and Students. The surveys were posted within the first week of July to all groups and were available through August 31, 2021.
Data Analysis
Comparisons between burnout and productivity were conducted using the general linear model and are reported using a regression framework to demonstrate comparisons between continuous variables after accounting for demographic (e.g., age and gender) and contextual (e.g., employment setting, years in setting, role) information. We also conducted an analysis of productivity by setting and practitioner type, using the general linear model to allow the inclusion of covariates. Because the variables of interest are both dichotomous and polytomous, we report the results using an analysis of variance framework. Assumptions for all analyses were assessed by means of graphical methods, such as quantile–quantile (QQ) plots and residual plots. Because of the ordinal nature of our measures of ethical pressures, measures of relationship between productivity and ethical pressures were assessed using Spearman’s ρ.
Additionally, a check for univariate and multivariate outliers found that two datapoints were drastic outliers in one variable of interest (productivity). We calculated all analyses with and without these outliers. The outliers were removed because they negatively affected the assessments of normality and homoscedasticity, yet the interpretations of parameter estimates did not meaningfully change.
Results
Descriptive Productivity and Burnout Results
Respondents provided a specific numeric productivity standard or a range, depending on employer requirements. The average productivity numeric was 87.0% (n = 325; SD = 5.715; range = 50.0%–100%). Practitioners (n = 51) who reported a range for their productivity averaged a low of 83.6% and a high of 93%. The mean productivity standard for clinicians was 87.1% in SNFs and 85.4% in ALFs.
The majority of practitioners viewed their productivity standard as unreasonable (80%; n = 262) and indicated that an appropriate productivity standard would be, on average, 74.4% (n = 329). Most respondents (70%) reported that COVID-19 did not provide relief from productivity standards.
Data in Table 2 include the EE, DP, and PA means for this sample in comparison with normative data from the MBI manual. Our EE mean was 15 points higher, and the DP mean was over 3 points higher, than any of the MBI–HSS (MP) listed means. Consistency is evident between the means found in this study and those reported in the MBI manual for PA.
Burnout Dimension Respondent Means Compared With MBI–HSS Means
Note. Reproduction by special permission of the publisher, Mind Garden, Inc., www.mindgarden.com, from the Burnout Inventory Human Services Survey by Christina Maslach and S. E. Jackson. Copyright 1981 by Christina Maslach and Susan E. Jackson. Further reproduction is prohibited without the publisher’s written consent. MBI–HSS = Maslach Burnout Inventory Human Services Survey.
aIndicates the sample size for this survey and sample sizes of comparative means provided by Maslach et al. (2018).
Table 3 depicts burnout profiles across practitioner roles. Of all practitioner respondents, 20.5% were “burned out,” 84.2% were “overextended,” 57.9% were “disengaged,” and 30.6% of participants classified as “ineffective.” Only 12.6% of participants were “engaged,” with favorable scores on all three domains. It was possible for each respondent to be represented in more than one dimension. For example, a significant number (35.5%) of respondents had extreme scores for EE and DP, but not PA. These respondents nearly missed a burnout classification by maintaining a sense of professional efficacy and accomplishment.
Burnout Profiles by Practitioner Role
Note. Profiles and scores are as follows: Overextended = extreme scores (≥27) on the Emotional Exhaustion (EE) scale; Disengaged = extreme scores (≥10) on the Depersonalization (DP) scale; Ineffective = extreme scores (≤33) on the Personal Accomplishment (PA) scale; Burned Out = extreme scores in all three dimensions; Engaged = favorable scores in all three dimensions.
aTotal indicates combined n and average percentage of all practitioners who met the profile criteria.
Burnout and Productivity in Context
We used general linear models to assess the relationship between burnout subscales (EE, DP, and PA) and productivity, after accounting for demographic and contextual information. Residual and QQ plots showed no meaningful deviation from statistical assumptions for the DP and PA models, whereas the QQ plot for the EE model indicated a slight deviation from normality in the upper tail. In light of the robustness of regression to slight deviations in normality, we believed that it was appropriate to continue with the EE model.
After accounting for demographic and contextual information, we found statistically significant positive relationships between productivity and both EE (p < .01) and DP (p < .01), whereas there was a marginally statistically significant negative relationship between productivity and PA (p = .05). That is to say, higher productivity levels were associated with higher levels of EE and DP and lower levels of PA. See Table 4 for all parameter estimates of the relationship between productivity and each burnout subscale after accounting for demographic and contextual factors.
Regression Results With 95% Confidence Interval for the Three Dimensions of Burnout
Note. ALF = assisted living facility.
*p < .10. **p < .05. ***p < .01.
Productivity Analysis by Setting and Role
We used a general linear model to compare productivity rates across settings (i.e., SNF and ALF) and roles after accounting for other demographic and contextual factors. Results indicated that there was not a statistically significant difference in productivity between settings after accounting for other factors, F(1, 316) = 0.87, p = .35; whereas there was a statistically significant difference in productivity by role, F(4, 316) = 21.79, p < .01.
Further inspection revealed that the source of these differences in role was not between professions but between assistants and evaluating practitioners. All assistants (occupational therapy assistants and physical therapy assistants) reported significantly higher productivity rates than all therapists, with all ps < .01, whereas there were no significant differences reported for assistants or practitioners (p > .10).
Productivity and Perceived Ethical Pressures
Respondents indicated the frequency with which they experienced perceived pressure to participate in unethical behaviors. The percentage of practitioners who reported experiencing pressures frequently, consistently, weekly or daily was highest for pressure to keep patients on caseload longer (67.4%), increasing session length (54.5%), and billing for patients who refuse therapy (55.1%). Pressure to change clock time (33.5%), falsify documentation (28.7%) and change billing codes (17.3%) were reported at high levels for fewer respondents. Nine respondents (2.5%) indicated that they never perceived pressures to participate in any of the unethical behaviors listed.
We used Spearman’s ρ to determine the relationship between productivity and ethical pressures. Small statistically significant relationships were found between productivity and falsifying patient documentation (ρ = .12, p = .03), changing clock-in or clock-out time (ρ = .14, p = .01), keeping patients on caseload longer (ρ = .18, p < .01), increasing session length (ρ = .12, p = .03), and finding ways to bill for patients who refuse therapy (ρ = .16, p < .01). The only ethical pressure that was not statistically significantly related to productivity was the pressure to change billing codes (ρ = .06, p = .30).
Discussion
This study, for practitioners working in SNFs and ALFs, is one of the first to describe the contextual experience of productivity, burnout, and ethical pressures. Our findings suggest that as productivity standards increase, a practitioner experiences more EE and DP as well as lower feelings of PA. Maslach and Leiter (2022) have indicated that at any one time, a work environment can expect the overextended profile in 15% to 20% of the workforce, the disengaged profile in 20% of the workforce, and the ineffective profile in 15% to 20% of the workforce. Our findings of 84.2% overextended, 57.9% disengaged, and 30.6% ineffective practitioners suggest a different reality for practitioners in the current geriatric health care environment. There appears to be a call for action in our professions when 20% of practitioners are characterized as burned out and another 67% are demonstrating extreme exhaustion, disengagement, or low professional efficacy. Those managing levels of stress and change without significant incident (12.6%) characterized themselves as engaged. A more typical engagement profile is approximately 30% of a working sample (Maslach & Leiter, 2022). The high levels of EE seen in this study are consistent with the findings of Zambo Anderson et al. (2015) and NASEM (2019), which associate burnout with those who have jobs that demand increased effort. The recommended productivity standard of 74% in this study is similar to the previously recommended 72.3% for therapists and 77.5% for assistants (NASEM, 2019).
Although Tammany et al. (2019) found that SNFs had the greatest prevalence of high productivity standards, our study found no significant difference in productivity between SNF and ALF settings. The statistically significant findings across roles, with a higher rate of productivity required from assistants than from therapists, are consistent with professional expectations in the field.
Reported pressures to consider unethical behaviors in the context of productivity standards are concerning. The statistically significant relationships between productivity and perceived pressures toward unethical behaviors represented in our data reinforce findings from earlier studies (Cantu, 2019; Tammany et al., 2019; Warren & Sampson, 2020) where respondents felt that productivity requirements pressured them toward unethical choices.
Implications for Occupational Therapy Practice
Considering the detrimental impact of burnout (Bostock, 2019, Maslach, 2018; Salyers et al., 2017), as well as the quality of life of practitioners and our ability to retain and recruit practitioners, mitigating actions are warranted. When workload is a major contributor leading to exhaustion and burnout, as we have seen in our sample, there are multiple areas to address: ▪ Maximize recovery from exhaustion by building strength and resilience. This may include enhancing mindfulness and/or coping strategies and utilizing stress management. Supportive personal relationships; spiritual practices; re-examining work attitudes and life philosophy; and self-care practices such as exercise, hobbies, adequate sleep, and proper nutrition (De Hert, 2020; Munhoz et al., 2020) may all have potential benefits. ▪ Consider rebalancing resources and demands, with a focus on “working smarter, not harder” (Maslach & Leiter, 2022), primarily by the employer, but it can also be proposed or initiated by the employee. ▪ Exhaustion may be moderated by keeping clear boundaries between work and home (Maslach & Leiter, 2022). ▪ Instill students, new graduates, and clinicians with new ways to reconfigure their passion for their profession. ▪ Promote advocacy efforts that heighten the respect and professionalism of clinicians’ careers while also protecting the best interests of patients.
A Change in Meaning
NASEM (2019), in its report “Taking Action Against Clinician Burnout: A Systems Approach to Professional Well-Being” calls for immediate action to address clinician burnout. Among their recommendations is to apply human-centered design in developing payment and performance policies and standards to lessen adverse effects on clinicians. Changing the definition of productivity, a standard created by employers to enhance efficiency, should be viewed as an achievable goal. Redefining productivity can keep employees accountable while recognizing essential but nonbillable tasks. It may be “productive” for a therapist to complete a brief self-care session, take time to reflect on patient successes and outcomes, or be able to pause and assist a colleague. Job tasks and demands that are necessary but not currently within productivity definitions should be reconsidered, including documentation, interprofessional collaboration, chart reviews, completing training, supervision, and cleaning equipment (NASEM, 2019). These small steps could be something that employers can support, and even encourage, to the benefit of all constituents.
Productivity standards affect one’s ability to make their own decisions, as evidenced by the pressures experienced by our participants. Workers are 43% less likely to have burnout when autonomy in the workplace exists (Pendell, 2018). Employees can gain some control over their situation by asking for slight modifications and lessening exposure to situations that deplete their energy (Maslach & Leiter, 2022).
The Work Institute Retention Report (2020) recommends that employers challenge employees to develop their skills and that they provide opportunities for growth and feelings of contribution. Employers need to create a work community, provide adequate compensation, and use gestures to relay appreciation (Work Institute Retention Report, 2020).
Although 62.2% of practitioners indicated that they have taken steps to advocate for conditions that would better support practice and ethical decision making, a large percentage of practitioners experience elements of burnout and do not speak out. All practitioners are well versed in their ethical codes, professional communication, and advocacy and should not hesitate to represent their best interests alongside those of their clients.
Limitations
A disproportionate number of respondents worked in SNFs, with only 6.8% in ALFs, limiting application to that setting. Selection bias may have been a factor, because recruitment materials clearly emphasized burnout as a topic. Conversely, individuals who are truly burned out may have neglected to participate because of having left professional social media or feeling overwhelmed.
Future Research
In further studies, inclusion criteria could be broadened to incorporate individuals who have recently left these fields. Research should also be conducted with companies that have adopted a different definition of productivity to determine how burnout is evident in those settings.
Conclusion
Practitioners working in SNFs and ALFs are reporting feelings of burnout and particularly high levels of emotional exhaustion, which are associated with high productivity standards and perceived pressures from employers. This occupational risk can cause them to have both mental and physical ailments (Munhoz et al., 2020). The professions of occupational, physical, and speech therapy set high ethical standards, because the protection of the rights of those we serve is of utmost importance. Employers must begin to recognize the impact of burnout and the productivity pressures that can result from attempting to cut costs. Changing the definition of productivity, having practitioners adopt measures of self-advocacy, incorporating a supportive work environment, and education on coping strategies may potentially decrease pressure for unethical practices, prevent therapist burnout, reduce practitioner turnover rates, and improve overall patient care.
