Physical Therapy Utilization Among Hospitalized Patients With Pediatric Acute Lymphoblastic Leukemia

Abstract

PURPOSE:

Patients with pediatric acute lymphoblastic leukemia (ALL) are at risk for impaired physical function from treatment. Early physical therapy (PT) may improve physical function and health in children with ALL, yet little is known about PT utilization in this population.

METHODS:

Leveraging the Premier Healthcare Database, we conducted a cohort study including participants hospitalized with ALL at age 0-21 years from January 1, 2010, through March 31, 2017. A generalized mixed linear model assessed sociodemographic and clinical variables associated with receiving PT within 1 year of first hospitalization.

RESULTS:

Among 5,488 pediatric ALL patients from 330 hospitals (median age 7 years, interquartile range = 4-14 years), only 27.2% overall and 58.9% with neuromuscular conditions received PT within a year of first ALL admission. In multivariable analysis, patients more likely to receive PT were age 10-14 years (odds ratio [OR] = 1.46; 95% CI, 1.20 to 1.76) or 15-21 years (OR = 1.66; 95% CI, 1.36 to 2.02) versus 0-4 years and Hispanic (OR = 1.27; 95% CI, 1.04 to 1.56) versus White. Patients less likely to receive PT were treated by a nonhematology/oncology pediatric (OR = 0.56; 95% CI, 0.46 to 0.70) or adult (OR = 0.50; 95% CI, 0.38 to 0.65) specialist versus a pediatric hematologist/oncologist and treated at a nonteaching hospital (OR = 0.53; 95% CI, 0.36 to 0.79) versus a teaching hospital.

CONCLUSION:

Only 27.2% of pediatric ALL patients overall and 58.9% with neuromuscular conditions receive inpatient PT within a year of first ALL admission. Interventions to increase inpatient PT services to pediatric ALL patients and address disparities in PT utilization may improve the physical function and long-term health of survivors.

INTRODUCTION

Survival outcomes in patients with pediatric acute lymphoblastic leukemia (ALL) are excellent with the 6-year overall survival exceeding 95%.¹ Yet, survivors of ALL remain at risk for late effects of their cancer treatment² and are particularly at risk for conditions that can affect motor function including peripheral neuropathy,^3,4 myopathy,⁵ bone toxicities,^5,6 fatigue,⁷ and obesity.⁸ Improving motor function in ALL survivors is important because it is associated with lower risk of falls^9,10 and may also improve school participation, physical activity, and health and quality of life in long-term survivors.^4,11,12

Multiple studies have suggested the potential benefit of physical therapy (PT) to improve motor function in pediatric ALL patients, even in those without known neuromuscular conditions.^13-18 In the Stoplight Program, a study of 135 patients with ALL who received a standard PT evaluation and tailored PT intervention at initiation of ALL therapy, those who completed the intervention had motor performance similar to healthy controls that persisted more than 2 years postintervention.^17,18 Marchese et al¹³ reported that 13 pediatric patients receiving ALL therapy who were randomly assigned to a PT program had improved lower extremity strength and joint range of motion compared with their 15 counterparts who received standard care only. Additionally, in a recent randomized interventional study by Saultier et al¹⁴ of 80 children with cancer, participants who received early versus delayed exercise intervention had improved mobility, flexibility, lower limb strength, balance, and endurance up to 1 year postintervention.

Despite the potential benefit of PT shown in clinical trials, little is known about routine utilization of PT in pediatric ALL patients, and there are no guidelines dictating receipt of this intervention. In one single-center retrospective study, only 30% of 35 pediatric ALL patients were referred to PT during any phase of treatment, but receipt of PT was not evaluated.¹⁹ Another recent single-center study reported that among 102 pediatric patients with hematologic malignancies, 36.3% received inpatient PT.²⁰ PT utilization has not been studied nationally, and demographic, geographic, and patient factors associated with receipt of this service are unknown. Patterns of inpatient PT utilization may be particularly informative given the high frequency of planned and unplanned hospitalization that occurs early in ALL therapy.²¹ A better understanding of the prevalence of PT utilization and demographic and hospital factors associated with receipt of PT is needed to determine which patients are receiving this service and to inform interventions to increase PT utilization overall and in at-risk individuals.

To address these knowledge deficits, we conducted a large retrospective cohort study to assess real-world inpatient PT utilization among pediatric ALL patients overall and in patients with known neuromuscular conditions and identify hospital and patient factors associated with receipt of PT.

METHODS

Data Source

We conducted a cohort study leveraging the Premier Healthcare Database (PHD; Premier Inc, Charlotte, NC).²² The PHD includes comprehensive demographic and clinical data from geographically diverse hospitals in the United States and captures approximately 25% of inpatient admissions across all age and payer groups.²² The database includes patient demographics, diagnoses, and detailed charge records with information on diagnostic tests, medications, procedures, and other inpatient services, with dates relative to admission. The Yale Human Investigation Committee determined that this study did not directly involve human participants.

Study Population

Pediatric patients (age 0-21 years) hospitalized with ALL from January 1, 2010, to March 31, 2017, were included in this study. ALL diagnoses were identified by International Classification of Diseases (ICD), ninth revision, code 204.0 and ICD, tenth revision, code C91.0.^23,24 The first ALL admission was used as a surrogate for date of diagnosis since it is standard practice for patients with ALL to receive inpatient care at the time of diagnosis.²⁵ All patients were followed for 1 year after their first ALL inpatient stay. Patients were excluded if they (1) received bone marrow transplant (n = 66) or (2) had traumatic diagnoses unrelated to their cancer diagnosis (n = 171; Fig 1). Traumatic diagnoses included fracture, dislocation, sprain, crush injury, motor vehicle accident, or other traumatic injuries.^23,24

FIG 1.

Flow diagram. ALL, acute lymphoblastic leukemia; PHD, Premier Healthcare Database.

Definition of Outcome

The primary outcome was ALL-related inpatient PT, defined as any PT event that occurred during an ALL inpatient hospitalization. PT was identified using charge codes. Our primary interest was receipt of PT within 1 year of the first ALL admission. We also assessed PT during the first ALL admission.

Variables of Interest

We examined patient and hospital characteristics that could influence receipt of PT.^26,27 Patient characteristics included age in years (0-4, 5-9, 10-14, and 15-21), sex, race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, or other/unknown), and insurance type (Medicaid, commercial, or other). Hospital characteristics included teaching status (yes or no), hospital size (large ≥ 500 beds, medium 200-499 beds, or small < 200 beds), and geographic region (South, Northeast, Midwest, or West). We also identified the type of attending physician during the first ALL inpatient stay that was further categorized as pediatric hematologist/oncologist, adult hematologist/oncologist, other pediatrician, or other adult physician.

Neuromuscular conditions that are commonly associated with treatment for pediatric ALL and can benefit from PT were also identified using ICD-9 and ICD-10 codes and included peripheral neuropathy, myopathy, and avascular necrosis.^13,28-31

Statistical Analysis

Frequencies of PT use were tabulated overall and by neuromuscular condition status. Pearson χ² tests and Student's t tests were used to compare baseline characteristics between patients with and without receipt of PT within 1 year of first admission. To account for clustering of patients within hospitals, we used generalized linear mixed models with a random effect term for the hospital to assess association of baseline patient and hospital characteristics with receiving PT. The model included sex, race/ethnicity, insurance, history of intensive care unit (ICU) stay, hospital size, hospital teaching status, hospital geographic location, and type of attending physician. Hospital was added as a random effect term in the models to adjust for unobserved hospital-level factors. Age at diagnosis as continuous variables was used to assess possible trends. All analyses were conducted with SAS (version 9.4, Cary, NC), and all tests were two-sided with an α-value of .05.

RESULTS

Participant Characteristics

Our study cohort consisted of 5,488 pediatric ALL patients from 330 PHD hospitals (Fig 1). More than half of the patients were male (57.7%) and non-Hispanic White (56.0%), with a median age at the first admission of 7 (interquartile range [IQR] 4-14) years (Table 1). Patients were most often treated at large hospitals (60.2%), teaching hospitals (69.4%), and by a pediatric hematologist/oncologist (60.7%) or an adult hematologist/oncologist (19.5%). The median hospital length of stay of first admission was 6 (IQR 3-11) days.

TABLE 1.

Characteristics of Participants Overall and Those Who Received and Did Not Receive Inpatient Physical Therapy Within a Year of Admission

PT Service Pattern

Of the 5,488 participants, 840 (15.3%) received PT during their first inpatient stay and 1,491 (27.2%) received inpatient PT within 1 year of their first ALL admission. Among 426 participants with documented neuromuscular conditions, 58.9% received PT within 1 year of first ALL admission (Fig 2). Among those with neuromuscular conditions, the prevalence of PT utilization was lowest in those with peripheral neuropathy, with only 55.7% receiving PT. The median length of stay for any admission with PT was 9 (IQR 4-18) days. The median time from admission to PT service was 4 (IQR 2-8) days with a median number of PT days per admission of 2 (IQR 1-4).

FIG 2.

(A) Proportion of patients with ALL (n = 5,488) who received PT (i) on first admission and (ii) within 1 year of first admission overall and (B) proportion of patients with neuromuscular conditions who received PT within 1 year of first admission including (i) any neuromuscular condition (n = 426), (ii) peripheral neuropathy (n = 334), (iii) myopathy (n = 35), and (iv) avascular necrosis (n = 74). ALL, acute lymphoblastic leukemia; PT, physical therapy.

Factors Associated With PT Service

Patients who received PT within 1 year of first admission were more likely to have a neuromuscular condition; to have a history of an ICU stay; be treated at a small hospital, a teaching hospital, or a hospital in the West; and be cared for by a pediatric hematologist/oncologist (all Ps < .01) than those who did not receive PT (Table 1). There was no difference in sex or race/ethnicity between the two groups.

In the generalized linear mixed model, compared with patients age 0-4 years, patients age 5-9 years were less likely to receive PT (odds ratio [OR] = 0.72; 95% CI, 0.60 to 0.85) while those age 10-14 years (OR = 1.46; 95% CI, 1.20 to 1.76) and 15-21 years (OR = 1.66; 95% CI, 1.36 to 2.02) were more likely to receive PT (Table 2). Patients who had a history of an ICU stay were more likely to receive PT than those who did not (OR = 2.78; 95% CI, 2.35 to 3.29), and Hispanic patients were more likely to receive PT than White patients (OR = 1.27; 95% CI, 1.04 to 1.56).

TABLE 2.

Association of Baseline Patient and Hospital Characteristics With Receipt of Physical Therapy Within 1 Year of First Acute Lymphoblastic Leukemia Admission^a

PT utilization also varied by hospital characteristics. Patients treated at nonteaching hospitals were less likely to receive PT than those treated at teaching hospitals (OR = 0.53; 95% CI, 0.36 to 0.79), although there was no difference in PT utilization by hospital size. Patients treated in the Midwest had a lower likelihood of receiving PT than those treated in the South (OR = 0.55; 95% CI, 0.32 to 0.93).

Finally, compared with patients treated by a pediatric hematologist/oncologist, patients treated by a nonhematology/oncology pediatric (OR = 0.56; 95% CI, 0.46 to 0.70) or adult (OR = 0.50; 95% CI, 0.38 to 0.65) specialist were less likely to receive PT.

DISCUSSION

To our knowledge, this study was the first to assess the national prevalence of inpatient PT utilization and factors associated with inpatient PT utilization in pediatric ALL patients. In this large, representative sample of 5,488 pediatric ALL patients, only 27.2% of patients received inpatient PT within a year of their first ALL hospitalization. Even among the patients with known neuromuscular conditions, only 58.9% received PT. We also found several factors associated with PT that can inform clinical care. Although older age, Hispanic ethnicity, and ICU stay increased the likelihood of receiving PT, age 5-9 years, admission to a nonteaching hospital, or treatment by a nonhematology/oncology attending physician lowered the likelihood of receiving PT.

A key finding of this study was that < 30% of pediatric ALL patients overall and only 58.9% of patients with known neuromuscular conditions received inpatient PT within a year of their diagnosis. The percentage of patients receiving inpatient PT in our study is lower than the previously reported 36.3% of 102 pediatric patients treated for hematologic malignancies in a single-center study.²⁰ That study also found that 76.4% of the 17 patients with severe peripheral neuropathy were referred to PT.²⁰ It is possible there was less frequent PT utilization in our study because this study was performed at a teaching hospital, which we found is associated with increased utilization of inpatient PT. Additionally, a prior single-center study of PT utilization in 35 pediatric ALL patients (age 1-19 years) found only 30% of patients were referred to PT during the course of their treatment, with the majority referred in the first year.¹⁹

Our findings have important clinical implications. Although the benefit of PT has been most frequently described in pediatric ALL patients with neuromuscular conditions, a growing body of literature also suggests the benefit of this service in improving health and physical function in patients with ALL even without known neuromuscular conditions.^13-18 We demonstrate in a large, geographically diverse, representative sample that the majority of patients with ALL do not receive this important service. Furthermore, only 7.8% of patients in our study had a neuromuscular condition, which is less than expected on the basis of previous studies of neuromuscular conditions in children with ALL,^3,29 suggesting that we may have only captured severe neuromuscular conditions and even more patients with neuromuscular conditions may be present in the overall study population who are not receiving PT. These findings highlight the need for interventions to promote PT during inpatient hospitalization since most patients do not currently receive this service.

The association of age with receipt of PT is another key finding of our study. We observed that patients age 5-9 years were less likely to receive PT than those age 0-4 years. It is possible that more attention is given to very young children because of the importance of promoting key developmental milestones such as crawling and walking.³² Nonetheless, 5- to 9- year-old children are still in need of this important service, especially since this is a time of school entry, and motor dysfunction can affect school performance.^33,34 Children of this age are also susceptible to other adverse outcomes associated with motor dysfunction including obesity and emotional distress.^4,8 In addition, we found older age groups were more likely to receive PT than those age 0-4 years. This may be partially explained by previous findings that older patients with ALL experience more severe neuromuscular comorbidities.³⁵ Recognizing age-related disparities in PT utilization can help clinicians target this service to individuals at risk.

Another interesting finding of our study is that patients treated by nonhematology/oncology attending physicians were less likely to receive inpatient PT than those treated by a pediatric hematology/oncology attending. This may be partially due to decreased comfort and familiarity of nonhematology/oncology physicians with the common comorbidities associated with ALL.^28,29 There are other benefits of patients with ALL receiving inpatient care from cancer specialists; a study of 7,275 children, adolescents, and young adults with newly diagnosed ALL from the California Cancer Registry found improved survival in patients receiving care at specialized cancer centers.²⁵ Our study expands on this by highlighting that another benefit of being treated by a cancer specialist may be receipt of PT. Future studies exploring other variations in care between patients with ALL treated by a hematology/oncology attending versus other physicians may help improve the quality of care for children with ALL.³⁶

We also found that PT utilization had minimal variation by geographic region. In unadjusted analyses, we found that patients who received PT were more likely to be treated in hospitals in the West than those who did not receive PT, similar to a national study of PT utilization among adult patients with vestibular disorders that found PT was more frequently provided to patients in the West.²⁷ However, in a generalized mixed linear model, patients treated at hospitals in the Midwest were less likely to receive PT than those treated at hospitals in the South, but there was no significant variation between PT utilization in other geographic regions. Although the reason for these differences in care is not clear, it does highlight a potential role for standardized guidelines to ensure this service is implemented across geographic locations.

This study should be interpreted in the context of potential limitations. It relied on hospital billing claims and may have underestimated the true number of PT services. Additionally, given the nature of the PHD, we were only able to follow patients within the same hospital system, and we were not able to determine if the same patient was seen across different hospital systems. This could have overestimated the number of patients with ALL and may have led to over or underestimation of the frequency of PT. This method still allowed the opportunity to use a large national sample to estimate PT utilization, resulting in valuable benchmarks. Finally, owing to the nature of the PHD, we were only able to assess inpatient PT services and could not determine if patients received PT in an outpatient setting. However, estimating the prevalence of inpatient PT utilization provides valuable insight into understanding the current practices during hospital stays in frequently hospitalized patients with ALL.

In conclusion, < 30% of pediatric ALL patients overall and only 58.9% of ALL patients with known neuromuscular conditions receive inpatient PT in the year after their first ALL admission, despite the potential benefit of PT in this population. Interventions aimed at promoting this service during ALL hospitalizations could minimize disparities in PT utilization and benefit the health of pediatric ALL patients.

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Provision of study materials or patients: Xiaomei Ma

Collection and assembly of data: Rong Wang

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Physical Therapy Utilization Among Hospitalized Patients With Pediatric Acute Lymphoblastic Leukemia

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).