Secondary Logo

Journal Logo


Therapy Workloads in Pediatric Health: Preliminary Findings and Relevance for Defining Practice

Long, Jason T. PhD; Neogi, Smriti PhD; Vidonish, William MBA; Badylak, Julie PT; Reder, Rebecca D. OTD, OTR/L

Author Information
doi: 10.1097/PEP.0000000000000665


Clinicians in the United States are asked to perform increasing amounts of patient care without an understanding by administrators of the nonbillable workload that accompanies time spent in patient care (eg, care coordination, care team communications, and professional development). There is no gold standard across practice settings to establish a threshold for what is considered “productive.” Current metrics do not allow for easy measurement of time spent doing necessary but nonbillable work, and as a result staff can become overloaded with direct patient care responsibilities and caseload expectations that are too large. This can lead to a host of problems, including burnout and work outside of normal hours, which can in turn lead to high turnover, safety risks associated with less time and attention paid to each task, and less time available for value-enabling tasks (eg, evidence-based practice research), which can ultimately affect the quality of care.

Literature is developing to describe the difference between direct patient care and overall workload in health care. The largest numbers of studies focus on nursing.1–4 The value of understanding the relationship between direct patient care load and overall workload is clear; caseload standards, staffing strategies, and revenue planning are directly linked to these measures.5 The small body of evidence available to describe similar measures in occupational (OT) and physical therapy (PT)6,7 does not accurately reflect changes in health care practice over the past 10 years.

Health care reimbursement models are shifting from volume to value.8 This transition from fee-for-service care to value-based care means that, eventually, profits will be tied to outcomes in patients (eg, achievement of goals and reduced readmissions) and not to the amount of care provided (eg, the number of patients that can be seen in a day). These changes require a renewed consideration for how therapists spend their time, as a large caseload will no longer be the hallmark of a “productive” therapist. To optimize therapeutic time, we need a means of accounting for it. However, because current productivity models place value on the number of billable units,9 measures of unbillable time are not easily available. Understanding is important in the complex relationship between the time commitments associated with patient care, professional development, and activities required by the institution (eg, cardiopulmonary resuscitation certification and safety training). Maintaining standards of care and professionalism while remaining financially solvent is not new10; however, while previous definitions of therapist productivity were derived strictly from billable time, the concept of productivity based on all activities required to support the provision of patient care requires a new metric and a redefinition of practice.

The definition of this new workload metric first requires an understanding of how tasks are prioritized and how time is allocated within the current model of care. This requires a consideration of tasks across the spectrum of what the current model considers “productive,” from billable time spent in patient care through nonbillable time spent in professional development. To understand this distribution of practice-related tasks in a standard therapist workday, our group sought to measure workload data in real time from therapists providing patient care in a pediatric rehabilitation setting. The purpose of this report is to detail findings from the therapy practices at a single pediatric institution, and to establish baseline data for a new workload metric. By more completely describing the distribution of therapists' time in the context of current productivity standards, we seek to lay a foundation from which new categorizations of work can be defined that are better aligned with a value-based model.


This study took place within the physical therapy and occupational therapy departments at Cincinnati Children's Hospital, Cincinnati, Ohio. The practice environment has 628 beds of inpatient, outpatient clinic services at the main hospital location and 10 satellite locations. In fiscal year 2014, the medical center reported over 30 000 inpatient admissions, 99 000 emergency department visits, and 31 000 surgical procedures. The study received exempt approval from the Institutional Review Board, with informed consent requirements waived due to study methods, which involved only the observation of “standard-of-care” sessions and the absence of personal identifiers.

At Cincinnati Children's Hospital, therapists serve in several subspecialty settings, including inpatient psychiatry, pain management, and the Heart Institute. All services are contained within the hospital system (eg, no contracted services with local community agencies). In both inpatient (IP) and outpatient (OP) settings, therapists are supported by unlicensed personnel who provide infrastructure support including cleaning, patient transport, and in-session aid, but who do not provide direct patient care or skilled therapy services. Therapists are provided with professional development opportunities in the form of internal learning symposiums twice per year and are given financial resources to pursue continuing education and professional certifications.


Definitions of workload and caseload were adopted as follows: workload refers to all activities required and performed by a health professional related to provision of patient services. Workload includes face-to-face direct services to patients, as well as many other activities necessary to support programs, implement best practices, and ensure compliance with legislative and professional standards.11Caseload refers to the number of clients being served by a health professional, through either direct or indirect service delivery options.11

This preliminary study was designed to address the following questions:

  • What portion of time do therapists spend doing activities that are patient-related versus time spent doing activities that are not patient-related, but that are done out of necessity or other priority?
  • What is the difference in median elapsed time in patient-related and non-patient-related work for different areas of care?

Task Definition and Data Collection

This study involved the measurement of time spent performing different tasks. Prior to the start of data collection, however, the study team needed to define the full spectrum of work performed by a therapist. Under the broad umbrella of work, the team identified the different activity types encountered by a therapist during a standard workweek:

  • Patient activity. Activity directly or indirectly related to patients (eg, scheduling, chart review, evaluation, care coordination, and patient-related travel).
  • Hospital activity. Activities not patient-related but related to the work of the hospital or health care institution (eg, meetings, mandatory training, departmental projects, and management/supervisory tasks).
  • Professional activity. Activities not patient-related, but related to professional development (eg, training/education, research, performance management, and mentoring).
  • Other activity. Activities not included in any other category, but still take place during daily practice (eg, waiting, non-patient-related travel, administrative, lunch, and technology troubleshooting).

Within each of these activities, the team then identified the variety of tasks typically encountered by a staff therapist over the course of a standard workweek. These tasks were compiled into the task definition list (TDL).

The TDL included practice-related tasks and tasks that are not practice-related but occur during the workday. The task list was developed with a level of detail to be comparable across clinical situations and clinical disciplines. Split-definition sets were developed to differentiate between the IP and OP care environments, but only when a universal definition was deemed inappropriate. As part of a larger multicenter project, each member of a 12-member consortium of pediatric therapy institutions independently created a TDL. These 12 TDLs were combined into a single list that was used for the present study. This final TDL consisted of 102 tasks (see Supplemental Digital Content 2, available at:

Data collection was a systematic observation and recording of tasks during the daily practice of staff therapists. During a 4-hour block for each therapist observed, the time spent performing each task was recorded and annotated using WorkStudy+ 5 software (Quetech Ltd, Waterloo, Ontario, Canada) on Nexus 7 Android tablets (ASUSTeK Computer Inc, Taipei, Taiwan). A team of 4 data collectors conducted time studies using the definitions established in the TDL. These data collectors were rehabilitation technicians recruited from within the department. This was, by design, such that their role in supporting patient care meant that they were already familiar with the therapy environment and the tasks performed. Prior to the start of the time studies, all data collectors were trained in the methods of conducting time studies within therapy environments and the use of the hardware and software (3 days of direct observation experience and mentored data collection). Prior to the start of data collection, the data collectors' interrater reliability was assessed by having each collector independently review and score a videotaped clinical scenario. Interrater reliability was assessed calculating Fleiss' κ on observed tasks and intraclass correlation coefficient on the average observed time for each task among 4 raters.

Each in-field data collection session involved an individual data collector observing an individual therapist for 4 continuous hours. Therapists were randomly selected and given the option of opting out of participation. There were no opt-outs. Observations were randomly distributed among morning, midday, and afternoon time over the course of the week, and also between disciplines and care environments (Figure 1). The data were reviewed at the conclusion of 6 weeks.

Fig. 1.:
Data breakdown map of sessions by clinical discipline and care environment.

Data Analysis

The final dataset consisted of records detailing the time spent performing each of the 102 individual tasks on the TDL. In addition to dividing tasks into 1 of the 4 activity groups, each task was assigned a care type attribute (Table 1). Data collection sections were grouped based on clinical discipline and care environment. After assigning these attributes to both tasks and data collection sessions, several descriptive analyses were performed to characterize the data. Variability in the percentage of time spent in each type of care was assessed with box-and-whisker plots. Differences between clinical disciplines and care environments were assessed via nonparametric statistical analysis. The Wilcoxon signed rank test was used to compare median values, with P < .05 considered statistically significant.

TABLE 1 - Descriptions of Care Types
Direct patient care Activity that is performed with a specific patient. Patient interaction is a requirement. Interactions may be face to face, or directly via phone, e-mail, etc.
Examples include evaluation, treatment, patient scheduling, etc.
Indirect patient care Activity that is performed for a specific patient in the absence of patient interaction.
Examples include documentation, room set-up, coordination of care, etc, before or after a treatment session, when the patient is not present.
Nonpatient care Tasks that are not related to a specific patient or are not clinical in nature.
Examples include breaks, training and education, site meetings, etc.


Interrater Reliability

Reliability was assessed in terms of both task identification and task duration. Very good interrater reliability was reached in task identification with Fleiss' κ of 0.874 (z = 17.7, P = 0), which takes into account the amount of agreement that could be expected to occur through chance. The interrater reliability on task duration on tasks among raters was high. The intraclass correlation coefficient for mean observed time on tasks among raters was 0.93 with 95% confidence interval from 0.77 to 0.99.

In-Field Data Collection

The full dataset consisted of 93 observation sessions (Figure 1). While each data collection was scheduled for 240 minutes (4 hours), the median duration of a session was 234.5 minutes (range: 34.4-276.4 minutes). This range of findings reflects the reality of a typical day spent in patient care, from a day that ends early due to a high number of cancellations to a day that requires additional time due to care coordination for a complex patient.

The percentage of time spent in each type of care was analyzed in the context of care environment and clinical discipline (Figure 2 and Table 2). The percentage of time spent in direct patient care ranged from 35% (OT-OP-direct patient care) to 49% (PT-OP-direct patient care), and the percentage of time spent in indirect patient care ranged from 32% (OT-IP-indirect patient care) to 39% (OT-OP-indirect patient care). The percentage of time spent in nonpatient care (NPC) ranged from 18% (PT-OP-NPC) to 26% (OT-OP-NPC).

Fig. 2.:
Percentage of time spent in each type of care (direct patient care, indirect patient care, and nonpatient care), stratified by clinical discipline (occupational therapy and physical therapy) and care environment (inpatient and outpatient).
TABLE 2 - Average Percent Time Spent in Each Care Type, Stratified by Clinical Discipline and Care Environment
Care Type Discipline Care Environment Sessions, n Mean (95% CI)
Direct patient care OT Inpatient 19 35.0 (24.7-45.3)
Outpatient 24 43.8 (33.9-53.8)
PT Inpatient 19 45.5 (36.4-54.7)
Outpatient 31 49.0 (40.6-57.4)
Indirect patient care OT Inpatient 19 39.0 (29.1-48.9)
Outpatient 24 32.6 (26.5-38.7)
PT Inpatient 19 32.4 (26.2-38.7)
Outpatient 31 32.5 (27.4-37.6)
Direct + indirect patient care OT Inpatient 19 74.0 (62.9-85.0)
Outpatient 24 76.4 (66.1-86.8)
PT Inpatient 19 78.0 (68.4-87.5)
Outpatient 31 81.3 (74.8-88.3)
Nonpatient care OT Inpatient 19 26.0 (15.0-37.1)
Outpatient 24 23.6 (13.2-33.9)
PT Inpatient 19 22.0 (12.5-31.6)
Outpatient 31 18.5 (11.7-25.2)
Abbreviations: CI, confidence interval; OT, occupational therapy; PT, physical therapy.

Figure 3 depicts the variability in the percentage of time spent in each type of care. For each care type, these plots provide a standardized display of the data distribution, including median value, overall range of values (whiskers), interquartile ranges (IQRs; boxes) representing the range in which 25% to 75% of values fall, and outlier values.

Fig. 3.:
Box-and-whisker plots of variation in direct patient care, indirect patient care, and nonpatient care as stratified by clinical discipline and care environment. Interquartile range indicated by shaded box. Median value indicated by horizontal line within box. Outlier values indicated by asterisks (*).

There was minimal difference between clinical disciplines and care environments (Table 3). No comparison was statistically significant.

TABLE 3 - Comparison of Direct Patient Care, Indirect Patient Care, and Nonpatient Care Values Between Clinical Disciplines and Care Environments
Comparison Difference (95% CI) P
PT-OP-DPC vs PT-IP-DPC −3.45% (−17.14% to 9.65%) .58
PT-OP-IPC vs PT-IP-IPC 0.09% (−5.35% to 7.99%) .63
PT-OP-NPC vs PT-IP-NPC 3.55% (−4.84% to 10.86%) .29
PT-OP-PC vs PT-IP-PC −3.55% (−10.86% to 4.85%) .29
OT-OP-DPC vs PT-OP-DPC −5.18% (−17.41% to 8.79%) .59
OT-OP-IPC vs PT-OP-IPC 0.08% (−5.53% to 9.48%) .58
OT-OP-NPC vs PT-OP-NPC 5.10% (−5.02% to 10.46%) .36
OT-OP-PC vs PT-OP-PC −5.11% (−10.46% to 5.02%) .36
OT-IP-DPC vs PT-IP-DPC −10.53% (−24.05% to 4.39%) .18
OT-IP-IPC vs PT-IP-IPC 6.52% (−6.51% to 16.57%) .35
OT-IP-NPC vs PT-IP-NPC 4.00% (−5.97% to 12.47%) .68
OT-IP-PC vs PT-IP-PC −4.01% (−12.47% to 5.97%) .66
OT-OP-DPC vs OT-IP-DPC −8.80% (−22.69% to 2.49%) .13
OT-OP-IPC vs OT-IP-IPC 6.35% (−6.78% to 15.55%) .41
OT-OP-NPC vs OT-IP-NPC 2.45% (−5.49% to 11.83%) .49
OT-OP-PC vs OT-IP-PC 2.45% (−0.69% to 31.85%) .08
Abbreviations: CI, confidence interval; DPC, direct patient care; IP, inpatient; IPC; indirect patient care; NPC, nonpatient care; OP, outpatient; OT, occupational therapy; PT, physical therapy.


In the current health care climate of what seems like constant change, the concept of clinician productivity is shifting. The productivity metric is evolving from a purely financial calculation to one involving many facets of patient care. Given this change, the role of the clinical provider in each of these areas must be accounted for and new methods of measuring this growing workload must be developed.

This study represents the first report of task-based time utilization in a pediatric therapeutic environment. The percentage of time spent in different types of care was consistent between outpatient PT, inpatient OT, and inpatient PT, with 42% to 49% of time spent in direct patient care and approximately 32% of time spent in indirect patient care. There was a shift away from this distribution pattern for outpatient OT, as observed time was split more evenly between each of the 3 care types. This distribution of care provided a clear opportunity for improving task delegation, with NPC leveraged to nonclinicians as much as practical. This hypothetical reassigning of tasks would allow an increased focus by therapists on direct and indirect patient care.

Throughout the different comparisons reported in this study, the variability between groups is worth noting. The variability addresses the ranges of time measured, as graphed in Figure 3. It is reasonable to question whether this variability is associated with either the task-based time study methods or the observers who collected the data. The detail of the 102-element task list, however, suggests that tasks could be captured at a high enough resolution to clarify the nature of the work performed (eg, differentiating patient care-based travel from professional development-based travel). Additionally, the training program completed by the observers, and subsequent findings of reliability from the test-retest sessions, suggests that this variability is not associated with the observers. The next most likely source of variability is the effect of the group. Time spent in direct patient care is most variable in the inpatient OT environment and least variable in the inpatient PT environment (Figure 3).

Nonparametric statistical analysis found no significant differences between either discipline or environment for any of the care types. While this suggests a level of equivalence among the 4 groups, the variation in percent times indicates that further study may be required to better understand some of the factors underlying task performance. Specifically, variation in the percentage of time spent in direct patient care was lowest in the inpatient PT group. While this group included only 19 observations, the comparably sized group of inpatient OT observations had higher variability. This relatively lower variability may have been related to underlying factors, such as staffing levels, patient population, treatment modality, or procedural standardization. Differences between care environments are also interesting to consider in the context of scheduling; inpatient therapists do have patients who are “scheduled” in the traditional sense of the word, but rather who are available and able to tolerate therapy. Outpatient therapists are reliant on both an efficient scheduling system and patient availability to maintain a high level of efficiency in using time for direct patient care.

Following this study, quality improvement (QI) projects were designed in both inpatient and outpatient environments to address issues that arose during data collection. Deficits in the patient scheduling process were identified as related to patient access, with an insufficient number of time slots available during several key scheduling periods. Using QI approaches, we identified errors in the process used by therapists to indicate available time and designed a system for continuously updating therapists' available time slots. This resulted in more patients scheduled and fewer unused time slots. We developed a system for targeted scheduling of time slots that remained open or became available (eg, due to cancellation), further reducing the queue of patients waiting to be scheduled. A second QI project focused on the delegation of tasks to nonlicensed staff in order to maximize therapist time spent in value-added and value-enabling tasks.

For all 4 groups, indirect patient care had the lowest variability based on IQR. In the inpatient PT group, the IQR was comparable to the NPC IQR, but for all other groups, the difference in variability was apparent. As indirect patient care-related tasks involved documentation and care coordination and given the strong influence of improvement science at this institution, this low variability may indicate the effective use (prior to this study taking place) of workflows and other solutions that allow these tasks to be performed efficiently.

With the exception of inpatient PT, the percent time spent in NPC had large variation across groups, as can be seen from the NPC IQR. NPC activities may be considered undesirable in a financially based measure of productivity, as they involve neither billable services nor the support of such services. However, NPC tasks, such as hospital-mandated safety training or collaborative development of best practice algorithms, are critical to the continued employment and professional growth of the clinical provider. Accounting for providers' need to carry out these NPC tasks will be necessary in the development of new workload guidelines. While many of these NPC tasks need to happen, it may be possible to optimize their timing so that patient care is minimally disrupted. Noncritical provider training that is appropriate for flexible scheduling could be scheduled during periods of low patient census. The results from this study may inform a diagnostic analysis to identify the cause of these NPC tasks and minimize their effect on patient care.

Data from this study have created a baseline from which new systems can be developed to address changes in the concept of caseload. They create a construct within which a group of patients and their associated care can be assessed from a task-specific perspective. In many settings, a therapist's caseload is considered as a particular volume of patients, possibly stratified by diagnosis or severity of impairment. For a particular outpatient therapist, this may create a current caseload of 30 patients, with a daily caseload of 5 to 6 patients (ie, in any given 8-hour day, 6 of the 30 patients are scheduled to be seen). If the concept of “individuals” is removed from this construct and replaced with “tasks,” a truer picture may emerge of what a clinician can realistically accomplish. For example, an especially complex patient may require additional levels of care coordination and family education, occupying more of the therapist's time spent in indirect patient care. In the current concept of daily caseload, this patient is 1 of the 6 on a given day. In the new framework, the additional time required for care coordination would be more appropriately quantified when considering what a therapist could manage.

While there is not a direct relationship between time spent in direct patient care and time considered as “productive” under the billable care model, the ratios of patient care time versus NPC time highlight opportunities for improvement. These opportunities were subsequently addressed via QI methodology. These QI projects were directed toward improvement opportunities in patient access, scheduling, task leveraging, and documentation efficiency. Reporting dashboards were introduced to managers and coordinators, allowing monthly updates on progress toward improvement. The results in this present report represent values from an “unimproved” therapy system against which subsequent improvements can be measured.

This study was limited by the reality of performing data collection in a working therapeutic care setting. A concern of observations studies is that persons behave differently when under observation, which may in turn affect study findings. We minimized the likelihood of this by identifying staff members who were familiar to the treating therapists to work as data collectors. We reasoned that people would be less likely to alter their behavior patterns if the observer were someone who would typically be present. Further, we collected observations over the course of a 4-hour session under the assumption that the people become more acclimated to being observed as the session progress.

Constraints of time, personnel, and resources limited the number of sessions that could be observed; members of the data collection team were temporarily reassigned from their roles in clinical care, but their absences were only manageable for a finite period. Additionally, this study did not control for repeat patient visits. A patient who was involved in multiple data collection sessions was treated as 2 different patients. Controlling for these repeat visits in future analyses may provide additional insight regarding the influence of patient-specific parameters on time spent in different tasks. Additionally, these results do not consider practice performed outside regular working hours, such as documentation completed at home.

Results from this study represent findings from a single institution over a single 6-week period. In the present delivery model for pediatric PT and OT services, only 42% to 49% of therapist time was spent in direct patient care. This means that more than 50% of time was nonbillable, which current revenue-based models of productivity would consider as nonproductive. This is not the case, however, as tasks performed during these hours foster things critical to the therapeutic care environment, such as care coordination, professional competency, and patient safety. This discrepancy between the billable care model and the reality of patient care further clarifies the need to redefine productivity, and to develop a new metric that more accurately reflects the many demands placed on a therapist's time. A more thorough study of the tasks that make up the indirect and NPC portions of a therapist's workload is warranted. A better understanding of these tasks may guide the development of a task management system that leverages NPC activities away from the therapist to a more appropriate role in the workplace. In the context of a value-based care system, this may warrant movement away from the current study's task categorizations and toward the actual task categories and subcategories. Additional attention is warranted toward the value placed on tasks by clinicians, as many tasks presently categorized as “value enabling” are in fact critical to the completion of “value-added” tasks.

Both of these future directions require standardized methods for collecting and interpreting workload data, and this study represents an initial step toward achieving that goal. Further study with additional institutions is warranted to understand how these values are affected by variation in other factors, including institutional culture, institutional infrastructure, patient demographics, and seasonality. In addition, follow-up analysis at this institution is warranted to understand the effects of the various QI projects implemented to address identified deficits. Expansion to other therapeutic disciplines is also warranted to study the repeatability and reproducibility of the minimal variation observed between PT and OT. The ultimate aim of this study is to create a universal standard for a reasonable workload, as well as an associated metric that directs and supports managerial decision-making. This measure will provide a comparison of actual versus targeted workload, assist with determining appropriate and safe staffing caseloads, assist in determining tasks that could be leveraged to nonlicensed personnel, and define standards that will allow for predictive staffing given fluctuations in population needs.

This is the first report of how pediatric physical and occupational therapists spend time in performing daily tasks. This study provides new information regarding time spent in different types of patient care, stratified by professional discipline and care environment. Results from this study will inform the development of a new productivity metric that better defines the practice of a pediatric therapist, and may guide caseload assignments and workload delegation, as well as defining standards for predictive staffing given fluctuations in population needs.


We are grateful to the following members of the planning and data collection teams at Cincinnati Children's Hospital Medical Center: Elizabeth Bauer, Kelly Dreyer, Michelle Kiger, Kyle Malblanc, Meredith Rudolf, Jillian Russell, Kristin Schulte, and Noelle Setters.

We are also grateful to our collaborators at the study partner sites who participated in the development of the final task list: All Children's/Johns Hopkins (St Petersburg, Florida), Ann & Robert H. Lurie Children's Hospital of Chicago, Children's Healthcare of Atlanta, Children's Hospital Colorado, Children's Hospital Los Angeles, Children's Hospital of Philadelphia, Children's of Alabama (Birmingham), Cleveland Clinic Children's Rehab Hospital, Connecticut Children's Medical Center, Cook Children's Medical Center (Fort Worth, Texas), and Primary Children's Medical Center (Salt Lake City).


1. Antinaho T, Kivinen T, Turunen H, Partanen P. Nurses' working time use—how value adding it is? J Nurs Manag. 2015;23(8):1094–1105.
2. Blay N, Duffield CM, Gallagher R, Roche M. A systematic review of time studies to assess the impact of patient transfers on nurse workload. Int J Nurs Pract. 2014;20(6):662–673.
3. Pelletier D, Duffield C. Work sampling: valuable methodology to define nursing practice patterns. Nurs Health Sci. 2003;5(1):31–38.
4. White DE, Jackson K, Besner J, Norris JM. The examination of nursing work through a role accountability framework. J Nurs Manag. 2015;23(5):604–612.
5. Schoo AM, R AB, Ridoutt L, Santos T. Workload capacity measures for estimating allied health staffing requirements. Aust Health Rev. 2008;32(3):548–558.
6. Cirrin F, Bird A, Biehl L, et al. Speech-language caseloads in the schools: a workload analysis approach to setting caseload standards. Semin Speech Lang. 2003;24(3):155–180.
7. Williams JI. Caseload and workload—a model for physiotherapy services. Hosp Health Serv Rev. 1986;82(3):120–123.
8. VanLare JM, Conway PH. Value-based purchasing—national programs to move from volume to value. N Engl J Med. 2012;367(4):292–295.
9. Kray KW. How did billable come to equal productivity, and what do we do now? AOTA Administration & Management Special Interest Section Quarterly. Vol 30. Bethesda, MD: The American Occupational Therapy Association, Inc; 2014:1–4.
10. Chew F, Kurfuerst S. Fostering clinical excellence while maintaining financial viability. AOTA Administration & Management Special Interest Section Quarterly. Vol 27. Bethesda, MD: The American Occupational Therapy Association, Inc; 2011:1–4.
11. Kirby KK. Hours per patient day: not the problem, nor the solution. Nurs Econ. 2015;33(1):64–66.

benchmarking; case management; occupational therapy; organization and administration; organizational innovation; patient care management; pediatrics; physical therapy

Supplemental Digital Content

© 2020 Academy of Pediatric Physical Therapy of the American Physical Therapy Association