Symptom Burden and Functional Gains in a Cancer Rehabilitation Unit

Abstract

Background/Aims

To determine if there is a relationship between patient symptoms and functional improvement on inpatient rehabilitation.

Methods

Retrospective review of medical records at an American tertiary referral-based cancer center of all patients admitted to an inpatient rehabilitation unit between 3/1/2013-5/20/2013. Main outcome measures included the Edmonton Symptom and Assessment Scale (ESAS) and Functional Independence Measure (FIM).

Findings

The medical records for 71 unique cancer rehabilitation inpatients were analyzed. Statistical analysis of total admission ESAS on total FIM change found no significant relationships. The symptom burden of the patients was mild. Patients demonstrated statistically significant improvements in function and symptoms during inpatient rehabilitation. The mean change in total FIM and total ESAS were an increase of 19.20 and decrease of 7.41 respectively. Statistically significant changes occurred in fatigue, sleep, pain, and anxiety.

Conclusion

Both symptom and functional scores improved significantly during inpatient rehabilitation. However, no significant relationships were found between symptoms at admission and improvement in FIM.

Introduction

Many cancer patients suffer from a variety of symptoms due to the disease and its treatments.¹ Multiple studies have shown that higher functioning cancer patients tend to have less symptoms.^2,3 Lin et al. found a strong relationship with pain, fatigue, disturbed sleep, distress and the Karnofsky Performance Scale in cancer patients.⁴ Hung et al. found a relationship between function and fatigue in non-small cell lung cancer survivors using the Karnofsky Self Report Performance Scale and the Brief Fatigue Inventory.⁵

This relationship has led to studies on physical activity’s impact on symptoms. Increasing physical activity through exercise programs has been shown in multiple studies to reduce cancer symptoms such as fatigue^6-11, cachexia¹², and pain.^13-21 Studies have demonstrated improvement in symptoms and or quality of life during inpatient rehabilitation of several populations including patients with multiple sclerosis,^22,23 chronic obstructive pulmonary disease,²⁴ lung transplant patients,²⁵ cystic fibrosis,²⁶ stroke,²⁷ general rehabilitation²⁸ and cancer.^29-31 These prior studies utilized a variety of instruments to assess symptoms and function.

While there has been published research regarding associations between symptoms and function/activity, research on the relationship of symptom severity and improvement/change in function has not been performed in cancer patients. One could postulate that symptoms such as fatigue, depression, and well-being could have an effect on rehabilitation participation and effectiveness. Could cancer symptom severity be predictive of functional improvement on inpatient rehabilitation? There have been prior analyses regarding the relationship of symptoms and functional improvement in stroke rehabilitation inpatients with no clear relationships found.^{32, 33}

The primary objective of this study is to determine if symptom burden at admission to inpatient rehabilitation is predictive of functional improvement in cancer rehabilitation inpatients. If such a relationship existed, it could help consulting physiatrists determine inpatient rehabilitation potential.

METHODS

Subjects

This retrospective study included all patients admitted to inpatient rehabilitation at an American tertiary referral based cancer center from 3/1/2013 through 5/20/2013. If a patient was admitted more than once, only one randomly selected admission would be included. Patients who had incomplete symptom or functional records were excluded.

Procedure

Our rehabilitation inpatients underwent the standard American acute rehabilitation facility therapy of 3 hours/weekday of a combination of physical therapy (PT), occupational therapy (OT), speech therapy (when appropriate) and/or group therapy. Every patient received at least one hour of PT and one hour of OT every weekday. The third weekday hour was either more individual PT/OT, group therapy or speech therapy, depending on the patient’s needs. Therapy on Saturdays consisted of approximately 1 hour of individual (PT or OT) or group therapy. There was no therapy given on Sundays.

Institutional Review Board (IRB) approval was obtained. A waiver of informed consent in compliance with federal and institutional guideline was granted by the IRB. Patients during this time period were analyzed because Edmonton Symptom and Assessment Scores (ESAS) were obtained during this time period. Physiatrists reviewed the medical records and collected the data. Collected data included the 12-item Edmonton Symptom Assessment Scores – Financial/Spiritual (ESAS-FS), functional information, demographic information, medical information, and laboratory values. Functional information included Functional Independence Measure (FIM) scores, reason for admission to rehabilitation, and length of stay on rehabilitation. Demographic information included age, race, sex, marital status, insurance type, and discharge disposition. Medical information included primary cancer type, length of stay prior to inpatient rehabilitation transfer, and if the patient returned to the primary acute care service. Laboratory values included white blood cell count, platelet count, prealbumin, albumin, and serum creatinine at the time of inpatient rehabilitation transfer.

ESAS subscores included the ESAS-Psychologic (sum of anxiety and depression scores) and ESAS-Physical (sum of pain, dyspnea, appetite, nausea, fatigue, and drowsiness) were calculated. FIM subscores included FIM-Activities of Daily Living (sum of eating, grooming, bathing, upper extremity dressing, lower extremity dressing and toileting), FIM-Mobility (sum of bed/chair transfers, toilet transfers, tub/shower transfers, mobility and stairs) and FIM-Cognition (sum of comprehension, expression, social interaction, problem solving and memory) were calculated.

Dependent-samples t-tests were conducted to compare the admission subscales and total ESAS score means paired to their discharge scores at a 95% confidence level. The same procedure was conducted on the FIM subscales and total scores to their discharge scores. In both analyses a specific, directional difference was hypothesized that due to patient improvement, symptoms would decrease and function would increase.

The third analysis directly tested the hypothesis that functional improvement could be predicted from admission symptoms. The hypothesized relationship between symptom severity and functional improvement were tested by examining the bivariate fit parameters of the total change in FIM by ESAS-psychological, ESAS-physical, and ESAS-total at the level of 95% confidence.

RESULTS

95 inpatient rehabilitation admissions occurred during the study time period (3/1/2013-5/20/2013). 23 patients with incomplete functional and/or symptom data were excluded. One patient was admitted twice to inpatient rehabilitation, therefore, only 1 of this patient’s admissions was analyzed. In total, 71 unique patient admissions were tested for this study.

Table 1 displays demographic information of the study group. The median age was 63 years (Standard Deviation (SD)=13.3). The primary cancer types, along with their frequencies (and percentages) are listed in Table 2.

Table 1. Categorical Demographic Variables.

Category	Frequency (Percentage)
Age
Below 65	43 (60.5)
Above 65	28 (39.4)
Race
White	49 (69.0)
Black	10 (14.0)
Hispanic	8 (11.2)
Asian	4 (5.6)
Sex
Female	39 (54.9)
Male	32 (45.0)
Marital status
Married	52 (73.2)
Divorced	8 (11.2)
Single	7 (9.85)
Widowed	4 (5.63)
Payer source
Private ins.	35 (49.2)
Medicare	35 (49.2)
Self-pay	1 (1.4)
Discharge Destination
Home	60 (84.5)
Acute	7 (9.8)
Skilled Nursing Facility	4 (5.6)

Table 2. Primary Cancer Types.

Cancer Type	Frequency (Percentage)
Primary CNS Lesions	24 (33.8)
Glioblastoma Multiforme	13
Anaplastic Astrocytoma	3
Medulloblastoma	2
Oligodendroglioma	1
Rheumatologic Brain Lesion	1
Pineal Tumor	1
Cervical Ependymoma	1
Cavernoma	1
Neurofibromatosis	1
Liquid Tumors	13 (18.3)
Multiple Myeloma	3
Leukemia	5
Lymphoma	5
Sarcomas	10 (14.0)
Chondrosarcoma	4
Liposarcoma	2
Fibrous Histiocytoma	1
Ewing’s Sarcoma	1
Pigmented Villonodular Synovitis	1
Hemangiopericytoma	1
Dermatologic	4 (5.6)
Melanoma	2
Dermatologic SCC	2
Head & Neck	2 (2.8)
Sinus SCC	1
Oral SCC	1
Genitourinary	4 (5.6)
Renal Cell Carcinoma	2
Prostate Cancer	2
Other Cancers	14 (19.7)
Rectal Cancer	1
Pheochromocytoma	1
Breast Cancer	6
Lung Cancer	6

CNS, Central Nervous System; SCC, Squamous Cell Carcinoma

Median white blood cell, platelet, and creatinine levels at rehabilitation admission were within normal limits. Median prealbumin and albumin were below clinically accepted limits at 18.1 and 3.0 respectively.

Patients did demonstrate statistically significant improvements in their ESAS scores. The median total ESAS score was 30 (M = 29.25, SD = 15.59) at admission and 19 (M = 21.84, SD = 15.16) at discharge. Total admission ESAS also was aligned with a significant proportion of variance in discharge scores, R² = .1289, F(1. 62) = 9.175, p < .0036. Table 3 lists ESAS scores at admission and discharge. Individual patients exhibited severe scores on occasional subscales but no patients exhibited a severe total ESAS score either at admission or discharge.

Table 3. Admission and Discharge ESAS Scores.

	Admission		Severe	Discharge		Severe
ESAS Score Item	M	SD	f	Freq	SD	f	df	t
Pain	3.25	2.77	8	2.26	2.56	2	64	−3.24**
Fatigue	4.24	2.55	8	2.66	2.59	3	63	−5.13**
Nausea	0.57	1.83	2	0.75	1.75	1	64	0.53
Depression	1.27	2.29	2	0.87	1.89	1	64	−1.38
Anxiety	1.64	2.34	3	0.81	1.59	0	64	−2.65*
Drowsiness	1.97	2.54	3	1.61	2.37	2	64	−1.36
Appetite	3.83	2.99	9	3.36	2.61	6	64	−1.52
Wellbeing	3.60	2.43	5	3.10	2.48	4	64	−1.56
Shortness of Breath	1.25	1.91	0	1.27	2.03	1	50	−0.44
Sleep	4.08	3.11	14	2.87	2.61	3	50	−2.36*
Financial	2.02	2.79	5	1.46	2.32	2	50	−1.20
Spiritual	1.19	2.14	2	0.76	1.62	1	50	0.06
Total	29.25	15.59	0	21.84	15.16	0	63	−3.75**

^*p < .05.

^**p < .001.

M=mean; SD=standard deviation; f = frequency; df=degrees of freedom; t=t-value Severe individual ESAS scores are considered to be scores of 7-10. A severe total ESAS score is considered to be 70-100.

The median length of stay before transfer to the inpatient rehabilitation unit was 10 days (SD=12.7). The median length of stay on inpatient rehabilitation was 11 days (SD=5.2). Patients with a longer length of stay also had a larger change in FIM (p=0.0003). Patients with a higher admission total FIM score tended to have a higher total discharge FIM score (b=0.355, t(62) = 3.03, p < .0036). Additionally, admission FIM scores significantly predicted discharge scores. Table 4 demonstrates changes in FIM from admission to discharge. Statistically significant changes were found in Total FIM scores and FIM subscales.

Table 4. Admission and Discharge FIM Scores.

	Admission		Discharge		Δ
FIM Subscale	M	SD	M	SD	M	SD	t (68)
ADL	33.08	6.36	42.20	6.40	9.15	5.25	14.48**
Mobility	17.36	4.57	24.42	4.50	7.02	3.97	14.54**
Cognition	30.97	5.25	31.57	4.35	0.65	1.96	2.60*
Total	88.54	14.34	107.79	14.90	19.2	9.93	16.06**

^*p < .05.

^**p < .001.

M=mean; SD=standard deviation; t = t-value

While significant improvements in cancer symptoms and functional improvement were found, further analysis of the relationship between ESAS and FIM found no significant results. Figure 1 demonstrates the correlation analysis of 1) Admission Total ESAS and Admission Total FIM, 2) Admission Total ESAS and Change in Total FIM, and 3) Absolute Change in Total ESAS and Change in Total FIM.

Figure 1. Regression Analysis of ESAS and FIM Scores.

DISCUSSION

The primary objective of our study was to analyze if a relationship existed between cancer inpatient rehabilitation admission symptom severity and functional improvement This is the first published study to do so in cancer patients specifically (a population that often suffers from a variety of severe symptoms) and the third to analyze this relationship in a rehabilitation setting on any patient population to our knowledge. Our study with a n=71 was unable to identify a statistically significant relationship.

Symptom improvements from admission to discharge were statistically significant. Multiple studies of cancer rehabilitation inpatients have shown similar results.³⁴ Heim et al. demonstrated improvement in global quality of life, physical well-being, and functionality from beginning to end of inpatient rehabilitation. Instruments used included the Functional Assessment of Chronic Illness Therapy Measurement System for Anemia/Fatigue, Hospital Anxiety and Depression Scale, and Multidimensional Fatigue Inventory.³⁵ Guo et al. obtained ESAS scores from 63 cancer rehabilitation inpatients and demonstrated a statistically significant improvement in anxiety, constipation, fatigue, pain, appetite, sense of well-being, and insomnia.²⁹ Functionally, patients improved significantly on inpatient rehabilitation which has also been demonstrated in prior studies with the American standard of 3 hours of therapy/day.^36,37 A relationship between a longer inpatient rehabilitation length of stay and increased change in total FIM has also been found.³⁸ The longer one receives intense rehabilitation the more improvement is to be expected. Our findings of longer stay is associated with increased functional improvement raises the possibility that there might be a “sweet spot” for rehabilitation as compared to cost. Portions of inpatient rehabilitation might be performed as an outpatient with the same functional improvement. More research is necessary to identify the length of stay to achieve optimal improvement.

With respect to our primary objective, Madden et al. studied health related quality of life and FIM scores of 116 Canadian post-stroke rehabilitation inpatients. Subscores and total score of the FIM along with Short Form-36 (SF-36) patient survey results were evaluated. They analyzed if there was a correlation between changes in SF-36 individual symptom scores and total symptom score with changes in FIM scores using Pearson correlation tests. Similar to our study, both outcomes improved significantly, but there was no correlation between the two.³² The SF-36 measures some symptoms including pain and vitality but also function. The ESAS measures symptoms only. In an inpatient and outpatient study of 223 Dutch post-stroke rehabilitation patients, fatigue versus functional scores were assessed at 6 months, 12 months, and 36 months post-stroke. Fatigue, basic ADL function, and advanced ADL function were measured using the Fatigue Severity Scale, Barthel Index, and Frenchay Activities Index respectively. The authors concluded that lower fatigue severity scale scores were not associated with longitudinal improvement in basic or higher activity of daily living scores.³³ The results of Madden et al., van de Port et al., and our study suggest that symptom severity does not affect functional improvement in rehabilitation.

There are limitations to our retrospective study. First, we only analyzed 71 patients. A preliminary analysis by a statistician hypothesized that a sample size of 60 should have been adequate to determine a relationship. Because of this, we felt 71 would have been able to reveal a relationship if one existed. Secondly, a proportion of patients had crucial data missing (23/94, 24.5%) that made analysis impossible. Third, this study consisted of a variety of primary cancer types, however, a more focused cancer population could have yielded more specific results. Last, the symptom burden of the majority of our cancer patients was quite mild. A severe individual ESAS symptom score is defined as 7-10. The percentage of patients with such scores never exceeded 25%. That being said, in our population, symptom severity was not a barrier to further functional improvement. Guo et al. also had similar mean ESAS scores ranging from 2.1 to 3.9 for the 7 worst symptoms in a group of cancer rehabilitation inpatients.²⁹ Using thousands of 9 item ESAS assessments (the first 9 items of the 12 item ESAS), a mean total score of 19.9 has been reported in cancer patients 6 months before death.³⁹ Total 9 item ESAS scores for our group were slightly worse at 21.98. Patients with higher symptom burdens have had lower function in prior studies.^2-5 Cancer patients who are accepted to inpatient rehabilitation are screened by a consulting physiatrist. Patients have to be able to tolerate 3 hours of therapy per day to qualify for acute inpatient rehabilitation. A relationship has been identified in prior studies between symptom burden and performance status.^4,5 The prequalification to tolerate 3 hours of therapy per day may have reduced the numbers of patients with high symptom burdens from the study.

There are a number of possible explanations for the lack of association between inpatient rehabilitation symptom burden and functional improvement. One possibility is that the physiatry team was able to successfully manage distressing symptoms such as pain, nausea, anorexia and fatigue during the rehabilitation admission, thereby allowing patients who had these symptoms to adhere to their intense rehabilitation. Our findings of significant improvement of overall symptoms and particularly pain and fatigue upon rehabilitation discharge are reassuring and suggest that the physiatry team may have successfully controlled these symptoms during the rehabilitation admission. Another possible explanation is that some symptoms/characteristics that are not measured with the ESAS could have impacted adherence to rehabilitation. These might include, for example, resilience or motivation. More research is needed to better understand the association of symptom burden, symptom relief and functional improvement in the cancer rehabilitation population.

CONCLUSIONS

The symptom burden and function of these cancer rehabilitation inpatients improved during the rehabilitation admission. No significant relationship was found between admission total ESAS scores and total FIM change.

Table 5. Bivariate fit of FIM total change by ESAS subscales and total.

Subscale	M	SD	Median	n	Parameter	SE
Psychological	2.92	4.01	1.0	71	−0.0734	0.299
Physical	15.2	14.5	18.0	70	−0.0472	0.145
Total	26.2	13.9	27	70	−0.0132	0.080

Note: None of the bivariate parameters were statistically significant.

M=mean; SD = standard deviation; SE = standard error.