Secondary Logo

Journal Logo


Safety of Symptom-Based Modification of Physical Therapy Interventions in Pediatric Oncology Patients With and Without Low Blood Counts

Gilchrist, Laura PT, PhD1,2; Tanner, Lynn R. PT3

Author Information
doi: 10.1097/01.REO.0000000000000042
  • Free


One major side effect of treating cancer with chemotherapy is bone marrow suppression, resulting in negative effect on circulating red blood cells (RBCs), white blood cells (WBCs), and platelets.1 Myelosuppression can be caused by direct action of chemotherapeutic drugs on the myeloid stem cells or by indirectly suppressing hematopoiesis. The effect of myelosuppression can be anemia, neutropenia, and/or thrombocytopenia. These low blood counts have the potential to lead to adverse events, especially when the body is placed under stressful conditions such as exercise.2–4 Some of the most concerning adverse events include bleeds due to thrombocytopenia, cardiac arrhythmias due to severe anemia, and infections leading to sepsis due to neutropenia.5 The APTA Academy of Acute Care Physical Therapy has provided guidance to physical therapists on the suitability of exercise and other interventions when blood values are low. The initial guidelines6 (Table 1) were published in 1995 and provided information on appropriate exercise for patients with low hemoglobin (Hgb), hematocrit, WBC, and platelet values. In 2011 and 2013, the Academy for Acute Care Physical Therapy modified these guidelines to allow for more activity at lower RBC and platelet counts (Table 1).7 The Academy also noted that the cause for low counts is a factor that physical therapists should consider when applying these guidelines. Childhood cancer patients undergo treatment that may last between 3 months to 3 years, depending on the cancer type, putting children at risk for having chronically low blood counts. Thus, if the guidelines for activity were enforced with this population, many would receive recommendations for no or light exercise over a period of months to years.

TABLE 1 - Summary of APTA Academy of Acute Care Physical Therapy Guidelines for Exercise
Hgb, g/dL Platelets, per mm3 WBC, per mm3
Original Updated Original Updated Original Updated
No exercise <8 <8, but allow essential daily activities <20 000 <10 000 <5 000 with fever ...
Light exercise 8-10 <8-10, light aerobics, light weights 20 000-50 000 10 000-20 000 >5000 ...
Resistive exercise >10 >8 >50 000 >20 000 >5000 (as tolerated) ...

Children undergoing treatment of cancer are at risk for decreased physical activity during treatment, and this decrement in physical activity may last well after treatment and into survivorship. A recent study demonstrated that adult survivors of childhood cancer were more likely to be insufficiently active than their siblings.8 Low levels of physical activity are related to some of the known late effects of childhood cancer, including obesity, diabetes, and cardiovascular events.9–12 Thus, interventions aimed at increasing physical activity both during and after treatment of childhood cancer are thought to be important for long-term health. In addition, children undergoing cancer treatment often develop deficits in balance control, motor coordination, and gait impairments that may be mitigated by intervention and is often the goal of physical therapy (PT) in this population.13–16 An association between better motor coordination in childhood and increased physical activity as adults has been demonstrated, implying that early intervention for motor deficits may be important for long-term health.17 An additional possible benefit of exercise is the potential to speed Hgb recovery, as has been shown in adult cancer patients.18 Thus, to attempt to improve long-term health of childhood cancer survivors, both rehabilitation and oncology practitioners at Children's Hospitals and Clinics of Minnesota have been encouraging appropriate physical activity and participation in PT interventions for children undergoing cancer treatment, even during times of low blood counts.

When the oncology rehabilitation program was initiated at Children's Hospitals and Clinics of Minnesota, they had extensive discussions with the oncology providers about the safety of exercise and other interventions used by physical therapists in children with chronically low blood values due to chemotherapy. The oncology providers wanted to increase activity and participation in PT interventions and because of the chronicity of the low blood counts set standards with the physical therapists to identify when to modify PT interventions and physical activity based on patient presentation and not specific blood values (Table 2). No specific algorithm was set up to stipulate what activity modifications should occur, and physical therapists were allowed to use their clinical judgment, based on their education and experience, to determine how to adapt their interventions. This discussion took place within the context of the specific oncology patients treated by this institution, namely, that (1) patients were treated both as inpatients and outpatients by a core group of physical therapists working collaboratively; (2) the largest populations of patients treated had a diagnosis of leukemia, brain tumors, lymphoma, or other solid tumors; and (3) allogeneic bone marrow transplants are not performed at this facility. Because the clinical practice at this site differed from the APTA guidelines for blood counts in place, the opportunity existed to investigate the safety of symptomatic assessment prior to initiation of PT interventions in a population with the potential for chronically low blood values. Thus, the specific aim of this retrospective chart review was to look at the number and type of critical incidences that occur in children and adolescents with low blood counts (Hgb, platelet, and WBC) due to cancer treatment and compare the occurrence of adverse events with times when PT interventions were done and laboratory values were normal.

TABLE 2 - Guiding Principles for Symptomatic-Based Physical Therapy in Pediatric Oncology
Treat all patients as if they have depressed immunity and ensure clean equipment and environment
Modify exercise when:
Fatigue, headache, or pallor indicate symptomatic anemia
Bruising, petechia, or reports of abnormal bleeding (nosebleeds, gums) indicate thrombocytopenia
Fever (single reading of 38.3°C or at least 38.0°C for 1 h) is present
Resting vital signs (heart rate, blood pressure, respiratory rate, oxygen saturation) are abnormal
Other symptoms—examples: nausea, presence of a new rash


A retrospective chart review was performed to assess the safety of using symptoms to guide PT interventions in pediatric oncology patients. The study was approved by the Institutional Review Board at Children's Hospitals and Clinics of Minnesota. Consent to use medical records for research purposes was confirmed for all subjects. The subjects were a sample of 150 consecutive children, aged 3 to 18 years, with a diagnosis of a non–central nervous system cancer who received PT interventions, including strengthening, stretching, and gait and balance training, while undergoing chemotherapy and other treatments at a single institution. Only PT visits at 2 months or more from initiation of chemotherapy were included in the analysis to ensure the chronicity of the low blood counts. Subjects were excluded if they did not meet the inclusion criteria, had a previously diagnosed cardiac disorder, or had undergone a recent surgical resection.

For each subject, up to 4 PT intervention sessions were identified in the electronic medical record and relevant medical and PT intervention information was abstracted. Only PT sessions within 24 hours of a CBC count were reviewed. Although a blood count was done for these patients, this information was most often not available to the therapist at the time of intervention (ie, blood drawn immediately prior to or after intervention session). The patient's Hgb, platelet, and WBC values were recorded once a qualifying (within 24 hours of a CBC count) PT session was identified, along with the type of interventions performed, the patient's subjective symptoms, and if intervention was modified as recorded by the therapist. The time spent in the session was not recorded but generally lasted between 30 and 45 minutes on the basis of therapist availability and patient tolerance. Blood values were classified according to the Academy for Acute Care guidelines (Table 1) that were published in 19956 and in use during the inception of the study. A second classification was also done using the updated laboratory values resources,7 which were published in March 2013. The electronic medical record was examined for 2 days following the PT visit for any adverse events that may have occurred. The areas of the chart reviewed included PT documentation, hematology/oncology provider notes, nursing notes, phone messages to the clinic, and emergency department notes. Adverse events were defined by the treating physical therapists and oncology providers at the institution as syncope, cardiac arrhythmia, muscle or joint bleeds, low blood pressure (based on hospital normal values by age), oxygen saturation less than 90%, shortness of breath, bradycardia or tachycardia (based on age-adjusted normal values), falls, pain, impaired skin integrity, fever (single reading of 38.3°C or at least 38.0°C for 1 hour), and patient subjective complaints.


Once the data were gathered, a χ2 analysis was used to look for associations between adverse events and low Hgb, platelet, and WBC values. For cases where χ2 cells were fewer than 5, a Fisher exact test of significance was used. Risk ratio was also calculated to determine the relative risk of an adverse event following PT intervention when subjects' blood values were low as compared with sessions when CBC values were in normal limits.


The final population consisted of 147 patients who received PT interventions between June 2009 and July 2013; chart reviews of 3 patients were omitted because of missing data. The group was composed primarily of children who were being treated for acute lymphoblastic leukemia (ALL), but lymphoma and other solid tumors (including Wilms tumor, Ewing's sarcoma, and neuroblastoma) were also represented. A total of 406 PT visits were recorded, with 83% (338/406) occurring as outpatient visits. Mean Hgb, platelet, and WBC counts were below normal during the study visits (Table 3). As is shown in Table 4, Hgb was normal in less than 25% of the visits for any diagnostic population, platelet count was normal during approximately 40% to 45% of the visits, and WBC count was normal in less than one-third of the PT encounters. PT interventions were modified in 35 of the 406 session (8.6%). Strengthening exercise occurred in 91% of PT sessions, with resistance training (ie, body weight, resistance bands, or manual resistance) occurring in 81% of those sessions where strengthening occurred. Jumping was incorporated into almost half of the PT sessions (47.8%). Balance training occurred in 69.5% of sessions, and stretching was incorporated into 51.5% of sessions.

TABLE 3 - Patient Demographics (N = 147)
Mean ± SD (Range)
Age, y 8.7 ± 4.1 (3-18)
Time in treatment, mo 6.7 ± 4.5 (2-36)
Hgb, g/dL 10.3 ± 1.7 (5.9-15.7)
Platelets, per mm3 230.7 ± 132.2 (3.9-826.0)
WBC, per mm3 3.4 ± 3.3 (0.2-46.2)
n (%)
Gender—male 71 (48.3)
ALL 99 (67.3)
Lymphoma 14 (9.5)
Solid tumor 33 (22.4)
AML 1 (0.7)

TABLE 4 - Percentage of Physical Therapy Sessions With Normal Blood Values by Diagnostic Group
Diagnosis Hgb Platelets WBC
ALL 23.4% (77/288) 45.5% (131/288) 12.2% (35/288)
Lymphoma 9.7% (3/31) 45.2% (14/31) 29.0% (9/31)
Other solid tumor 18.1% (15/83) 42.2% (35/83) 32.5% (27/83)

A total of 37 events occurred within 2 days of patients receiving PT (Table 5), for a cumulative incidence of 9.1%. The most common event was tachycardia (n = 21), followed by subjective complaints (n = 9) and development of fever (n = 4). The subjective complaints included headaches (n = 4), abdominal discomfort (n = 3), and fatigue (n = 2). In none of the cases of tachycardia was medical intervention initiated to specifically address the issue. The 2 most concerning adverse events noted during the study period occurred during the PT session. One was a fall within the PT session where no injury was reported, and the other was the onset of a headache with visual changes noted during PT treatment. The latter event did lead to imaging and medical follow-up, but symptoms resolved within 48 hours. Severe adverse events, including muscle and joint bleeds, cardiac arrhythmias, syncopal episodes, periods of low oxygen saturation, and infections leading to sepsis, were not found within 48 hours of PT intervention. Diagnostic group (ALL, lymphoma, solid tumor) and inpatient versus outpatient status were not associated with occurrence of adverse events (χ2 = 4.50, P = .21; χ2 = 0.17, P = .68, respectively). None of the PT intervention types (strengthening with or without external resistance, jumping, stretching, balance, endurance) were associated with an increased risk of adverse events.

Fig. 1.:
Incidence of adverse events versus no event within 48 hours of a physical therapy session based on Hgb or Plt level. Panels A and C include all events, whereas panel B omits incidents of tachycardia. Hgb indicates hemoglobin; Plt, platelet.
TABLE 5 - Adverse Event Type by Diagnostic Group
Event Type ALL Lymphoma OST Total
Tachycardia without intervention 16 2 3 21
Tachycardia with medical intervention 0 0 0 0
Fall 1 0 0 1
Pain complaint 1 0 0 1
Integument injury 0 1 0 1
Fever 1 2 1 4
Other subjective complaint 5 1 3 9

Examination of the events by classification on the 1995 blood value guidelines,6 in place at the time of study inception, yielded a significant difference in the number of events that occurred according to Hgb level (χ2 = 9.77, P = .002, ϕ = 0.16). When the group was divided into normal versus abnormal values, a risk ratio of 12.35 was obtained (95% confidence interval [CI], 1.67-91.33), indicating an increased risk of events in those with low Hgb levels. Since tachycardia is a known compensation for low Hgb levels and was not sufficiently concerning to initiate medical treatment, the analysis was rerun excluding tachycardia with no medical intervention. The removal of incidental tachycardia decreased the χ2 value to 3.05, which was nonsignificant (P = .08) and the risk ratio decreased to 5.12 (95% CI, 0.67-39.09). Platelet counts were rarely in the lower levels of less than 20 000/mm3 or 20 000 to 50 000/mm3 during therapy sessions and were not associated with a significant difference in incidences of adverse events (χ2 = 1.47, P = .23). Normal versus abnormal WBC counts were not associated with the occurrence of adverse events post-PT (χ2 = 0.49, P = .83).

Updated Guidelines

The guideline update in 2013 allows for resistance exercise as long as Hgb level is above 8 g/dL.7 Thus, we analyzed the incidence of adverse events when blood values were below 8 g/dL and at or above 8 g/dL (Figure 1). We found that there was an increased risk of adverse events in those with Hgb levels below 8 g/dL when incidental tachycardia was included (χ2 = 13.52, P < .001; RR = 4.18; 95% CI, 1.85-9.46); however, this was no longer significant when incidental findings of tachycardia were removed (χ2 = 1.24, P = .27; RR = 2.03; 95% CI, 0.56-7.42). In the updated guidelines, resistance exercise is allowed for patients with a platelet count above 20 000/mm3; thus, we analyzed the data using 20 000/mm3 platelet count as a divider. When platelets were below 20 000/mm3, a higher rate of adverse events did occur (χ2 = 11.08, P = .001; RR = 4.91; 95% CI, 1.76-13.67). In the 5 events that did occur in this population, 3 were incidences of incidental tachycardia, 1 was of fever, and 1 was complaint of a headache. Two of these events occurred in hospitalized patients and 3 occurred in outpatient treatment.


Overall, adverse events were rare in this population of pediatric cancer patients during and after receiving exercise interventions provided by physical therapists. This supports the literature that has found that exercise and other interventions used by physical therapists can be completed in a safe manner.19–21 Although we recorded 37 adverse events in the 2 days following 406 PT intervention sessions when using the symptom-based approach, 21 of these events were classified as incidental findings of tachycardia that received no specific medical intervention. Severe adverse events, including muscle and joint bleeds, cardiac arrhythmias, syncopal episodes, periods of low oxygen saturation, and infections leading to sepsis, were not found to occur within 48 hours of a PT intervention. Only 2 concerning events occurred during PT intervention sessions, a fall and a headache with visual changes, both of which resulted in no long-term health concerns. Given that Hgb levels were normal in less than 25% of PT sessions and that increasing heart rate via sympathetic activation is a normal adaptation to low Hgb levels,22,23 it was surprising that tachycardia was not found in more patients. In this study, tachycardia was typically noted when vital signs were taken as part of the intake for an oncology clinic appointment, or as part of the nursing assessment for inpatients, but not usually noted in the PT assessment. The most common remaining adverse events included onset of fever and subjective complaints of fatigue, pain, and headache. Given that most patients are immunocompromised for periods of time during chemotherapy treatment, it is not uncommon that infection and fever can develop. One estimate of the occurrence of fever with neutropenia in childhood cancer patients is 0.76 episodes per 30 days at risk,24 confirming that fever is a common occurrence in this immunocompromised population. Thus, our finding of less than 1% incidence of fever in the 48 hours post-PT is not surprising. It is likely that the inclusion of only those at least 2 months into treatment influenced this low rate. While a small number of adverse events did occur after PT intervention sessions, the side effects of myelosuppression may be responsible for many of these events, especially tachycardia and fever.

Patients with anemia and moderate levels of thrombocytopenia were able to tolerate multiple interventions performed as a part of their PT (modified by symptoms), including resistance training, without an increased risk of serious adverse event. Although platelet counts below 20 000/mm3 were rare in our population, this level was associated with an increase in adverse events but not necessarily bleeding-related events. It is likely that the low platelet counts were a sign of increased treatment toxicity and thus adverse events may or may not have occurred because of the PT intervention. From these data, we cannot assume a direct association between interventions performed as a part of PT and the number of adverse events with low platelet counts, but it does speak to the need to monitor those with low platelet counts for adverse events and other treatment toxicities.

Interventions were noted to be modified by the physical therapists in response to patient symptoms in only 8.6% of the sessions, although blood values were below normal limits during most visits. With the low number of incidents occurring and the lack of severe incidents overall, the safety of a symptomatic approach to this population is supported. As our patients were at least 2 months into treatment, the chronicity of the low blood counts is likely a factor in the ability of patients to tolerate exercise such as strengthening with resistance even during periods of low counts. By 2 months into treatment, patients and their families typically have a better appreciation of the signs and symptoms of anemia and thrombocytopenia and the body has a chance to adapt to chronic anemia. In acute anemia situations, the patients may be more symptomatic and less likely to be able to tolerate exercise. The current practice at the institution of study is to introduce PT into care (with instructions to stretch and keep moving) within the first few weeks after diagnosis but to begin more intensive interventions 1 to 2 months after initiation of cancer treatment. This allows the medical treatment to stabilize the cancer, families to adapt to the cancer diagnosis, and, for many patients, also to avoid a time of intensive steroid treatment. Thus, the study design matches the clinical practice most commonly followed at this institution. However, the safety of this same approach when patients are treated within the first few weeks after diagnosis is yet to be evaluated.

One factor in the safety of this approach is the familiarity of the providers with pediatric oncology populations. Many but not all of the PT sessions analyzed were provided by a core group of physical therapists whose clinical practice focuses on the pediatric oncology population. They have significant experience with this population (range, 3-10 years). Therapists with less experience and mentoring may not be able to replicate these outcomes. The inclusion of only PT sessions when a blood count was done within 24 hours may have influenced the results. Most often blood counts are done for ensuring readiness for treatment or to monitor treatment toxicities. Thus, including PT sessions only when there was a corresponding blood test may have increased the likelihood of adverse event reporting due to the tie to medical treatment. The utilization of one institution for this study also limits the ability to generalize the findings. In this institution, not all types of cancer patients and treatment possibilities were represented. At the study institution, allogeneic stem cell transplants and proton beam therapy are not done and there is a limited representation of patients with bone tumors. The sample included a large proportion of subjects with ALL, the most common form of childhood cancer, as well as lymphoma and other solid tumors. Thus, this approach should be especially cautiously considered in instances of stem cell transplant, bone tumor populations, and when the therapists are unfamiliar with treating the pediatric oncology population.

Given that many children and adolescents treated for cancer will have extended periods of time where they are at risk for low blood counts, therapists and other health care providers may be inclined to encourage low levels of exercise for safety purposes. However, given that the long-lasting and late effects of cancer treatment including obesity, diabetes, and cardiac disease may be in part offset by exercise, it is our hope that incorporating PT interventions and other forms of exercise into the patients treatment will lead to healthier survivors. This study supports that in many pediatric cancer patients, PT interventions can be safely incorporated into patient care and that the use of symptoms by physical therapists to guide clinical decision making can be a safe approach.


The authors thank Jessica Tice and Katie Peters for initiating this project when they were students at St Catherine University, as well as Katherine Wacker for her review of the manuscript. The authors also thank the rehabilitation staff, oncology providers, and the patients and families at Children's Hospitals and Clinics of Minnesota.


1. Goodman CC. Oncology. In: Goodman CC, Fuller KS, eds. Pathology: Implications for the Physical Therapist. 3rd ed. Philadelphia, PA: WB Saunders; 2009:chap 9.
2. Johansen KL. Anemia. In: Durstine JL, Moore GE, Painter PL, Roberts SO, eds. ACSM's Exercise Management for Persons With Chronic Diseases and Disabilities. 3rd ed. Champaign, IL: Human Kinetics; 2009:chap 32.
3. Moore GE, Lockard M. Bleeding and clotting disorders. In: Durstine JL, Moore GE, Painter PL, Roberts SO, eds. ACSM's Exercise Management for Persons With Chronic Diseases and Disabilities. 3rd ed. Champaign, IL: Human Kinetics; 2009:chap 33.
4. Schwartz AL. Cancer. In: Durstine JL, Moore GE, Painter PL, Roberts SO, eds. ACSM's Exercise Management for Persons With Chronic Diseases and Disabilities. 3rd ed. Champaign, IL: Human Kinetics; 2009:chap 27.
5. Irion GL, Goodman CC. Laboratory tests and values. In: Goodman CC, Fuller KS, eds. Pathology: Implications for the Physical Therapist. 3rd ed. Philadelphia, PA: WB Saunders; 2009:chap 40.
6. Garritan S, Jones P, Kornberg T, Parkin C. Laboratory values in the intensive care unit. Acute Care Perspect. 1995;3(4):7–11.
7. Acute Care Section-APTA Task Force on Lab Values. Lab values interpretation resources. Updated 2013. Accessed August 18, 2016.
8. Lown EA, Hijiya N, Zhang N, et al. Patterns and predictors of clustered risky health behaviors among adult survivors of childhood cancer: a report from the Childhood Cancer Survivor Study [published online ahead of print 2016]. Cancer. doi:10.1002/cncr.30106.
9. Veringa SJ, van Dulmen-den Broeder E, Kaspers GJ, Veening MA. Blood pressure and body composition in long-term survivors of childhood acute lymphoblastic leukemia. Pediatr Blood Cancer. 2012;58:278–282.
10. Mulrooney DA, Yeazel MW, Kawashima T, et al. Cardiac outcomes in a cohort of adult survivors of childhood and adolescent cancer: retrospective analysis of the childhood cancer survivor study cohort. BMJ. 2009;339:b4606.
11. Meacham LR, Sklar CA, Li S, et al. Diabetes mellitus in long-term survivors of childhood cancer. Increased risk associated with radiation therapy: a report for the childhood cancer survivor study. Arch Intern Med. 2009;169(15):1381–1388.
12. Steinberger J, Sinaiko AR, Kelly AS, et al. Cardiovascular risk and insulin resistance in childhood cancer survivors. J Pediatr. 2012;160(3):494–499.
13. Galea V, Wright MJ, Barr RD. Measurement of balance in survivors of acute lymphoblastic leukemia in childhood. Gait Posture. 2004;19(1):1–10.
14. Götte M, Kesting SV, Winter CC, Rosenbaum D, Boos J. Motor performance in children and adolescents with cancer at the end of acute treatment phase. Eur J Pediatr. 2015;174(6):791–799.
15. Leone M, Viret P, Bui HT, Laverdière C, Kalinova É, Comtois AS. Assessment of gross motor skills and phenotype profile in children 9-11 years of age in survivors of acute lymphoblastic leukemia. Pediatr Blood Cancer. 2014;61(1):46–52.
16. Gilchrist L, Tanner L. Gait patterns in children with cancer and vincristine neuropathy. Pediatr Phys Ther. 2016;28(1):16–22.
17. Smith L, Fisher A, Hamer M. Prospective association between objective measures of childhood motor coordination and sedentary behavior in adolescence and adulthood. Int J Behav Nutr Phys Act. 2015;12:75.
18. Courneya KS, Jones LW, Peddle CJ, et al. Effects of aerobic exercise training in anemia cancer patients receiving darbepoetin alfa: a randomized controlled trial. Oncologist. 2008;13:1012–1020.
19. Marchese VG, Chiarello LA, Lange BJ. Effects of physical therapy intervention for children with acute lymphoblastic leukemia. Pediatr Blood Cancer. 2004;42:127–133.
20. Ness KK, Esbenshade AJ, Friedman DL, et al. Feasibility and initial effectiveness of home exercise during maintenance therapy for childhood acute lymphoblastic leukemia. Pediatr Phys Ther. 2014;26:301–307.
21. San Juan AF, Fleck SJ, Chamorro-Vina C, et al. Effects of an intra-hospital exercise program intervention for children with leukemia. Med Sci Sports Exerc. 2007;31(1):13–21.
22. Shander A, Javidroozi M, Ozawa S, Hare GMT. What is really dangerous: anemia or transfusion? Br J Anesth. 2011;107(S1):i41–i59.
23. Glick G, Plauth WH Jr, Braunwald E. Role in the autonomic nervous system in the circulatory response to acutely induced anemia in unanesthesized dogs. J Clin Invest. 1964;43:2112–2124.
24. Castagnola E, Fontana V, Caviglia I, et al. A prospective study on the epidemiology of febrile episodes during chemotherapy-induced neutropenia in children with cancer or after hemopoietic stem cell transplantation. Clin Infect Dis. 2007;45(10):1296–1304.

anemia; cancer; exercise; thrombocytopenia

© 2017 Oncology Section, APTA.