Volume 38, Issue 5 , Pages 738-746, November 2009
Assessment of Symptoms Reported by 10- to 18-Year-Old Cancer Patients in Taiwan
Article Outline
Abstract
The purposes of this study were 1) to assess and describe the occurrence, frequency, severity, and distress of symptoms reported by Taiwanese pediatric cancer patients who were between 10 and 18 years of age, and 2) to use statistical analysis to determine whether the multiple dimensions (i.e., frequency, severity, or distress) of the Memorial Symptom Assessment Scale (MSAS) 10–18 alone can provide sufficient useful information for the assessment of symptoms that patients report as distressing. A total of 144 Taiwanese pediatric cancer patients and their mothers participated in this cross-sectional study. The frequency of symptoms for all patients ranged from 52% for “lack of energy” to 10% for “feeling nervous.” The most common symptoms (occurrence >40%) were “lack of energy,” “lack of appetite,” “feeling drowsy,” “sweating,” “worrying,” “nausea,” “dry mouth,” “pain,” and “lack of concentration.” Patients in the “on-treatment group” had more distressing symptoms than those in the “off-treatment group.” The severity and distress subscales did provide the most information for symptom assessment and were the two best subscales to represent the impact of symptoms on quality of life, fatigue, and internalizing behaviors. The findings of this study suggest that revising the current format of the MSAS 10–18 into three separate instruments (one for each of the subscales) might provide more accurate data for assessments. Such a modification would change the scoring system and provide for more accurate data analysis.
Key Words: Measurement, symptom distress, older children, adolescents, pediatric cancer, MSAS 10–18
Introduction
Pediatric oncology patients suffer from an array of symptoms, often beginning before their cancer is diagnosed and lasting not only throughout their treatment1, 2, 3 but also long after their treatment has ended4 and especially in the terminal phase.5 Although the treatment goal for pediatric oncology focuses primarily on cure, the symptoms of toxicity caused by the cancer treatment or the cancer itself are of major concern for patients as well as for their parents.2, 3 Severe distress from symptoms may delay scheduled treatments, decrease the effectiveness of treatment protocols, and delay the rehabilitation process. The alleviation of symptoms can have a great impact on patients' quality of life.6, 7 Thus, symptom management is a very important part of health care for pediatric oncology patients and their physicians, nurses, and other health care professionals. To have better symptom management, improved assessment of symptoms for pediatric oncology patients is needed.
Assessment of symptoms is typically performed by inquiring about either a single symptom or a select group of symptoms. Although these assessments do provide some clinically important information, they might not include symptoms that the child is actually experiencing and believes unimportant because no one is asking about them. Thus, current assessment methods may not be comprehensive enough to assess all the distressing symptoms a patient is experiencing. A more complete and broad-spectrum physical and psychological symptoms assessment may be more useful not only for research purposes but also for clinical practice. To date, the measurement tools available to assess multiple symptoms in pediatric oncology patients include the Therapy-Related Symptom Checklist8 and the Memorial Symptom Assessment Scale (MSAS)— MSAS 10–181 and the MSAS 7–12.9 Although there are some author/researcher developed measures reported in the literature, they have very limited, if any, psychometric testing.10 There are also reports, of qualitative descriptions of symptoms.11 These studies have provided some fundamental knowledge of symptom distress experienced by pediatric oncology patients.
Two methodological issues regarding symptom assessment still need to be clarified. The first is how to assess the multiple dimensions of symptoms, including occurrence, frequency, severity, and level of distress for individual patients.12 Portenoy et al.13 documented that distress alone can provide significant information about symptoms, but they also suggested that the combination of distress and frequency provides more information than distress alone. However, at this time there is no empirical evidence regarding the measurement of the multiple dimensions associated with symptoms in pediatric oncology patients. The second methodological issue is that there is no consensus on whether to use global symptoms of distress or to rely only on highly prevalent symptoms for clinical and research purposes. Assessment of global symptoms can provide information about the incidence of symptoms that patients experienced and capture clinically relevant information. Specific symptoms or a small group of highly prevalent symptoms collected may be enough for research purposes, but they may not provide enough information to be useful for clinical assessment.
Thus, the specific aims of this study were 1) to assess and describe the prevalence of symptoms in Taiwanese pediatric cancer patients aged 10–18 years and 2) to determine whether the subscales of the MSAS 10–18 (such as frequency, severity, or distress), together or individually, would provide enough useful information in the assessment of symptoms.
Methods
The findings presented in this study are from data collected as part of a longitudinal study to identify and test a model of the coping patterns of parents and their children who suffer from cancer. A study of using analytic procedures to derive clusters of symptoms experienced by pediatric oncology patients was published using the same data set.14 For the purpose of this analysis, only the baseline data of pediatric cancer patients aged 10–18 years and their mothers were used to describe the factors associated with the occurrence of symptoms.
Subjects
Detailed information about the recruitment of subjects and their characteristics has been described in an earlier published report.14 Data from 144 patients (84 male and 60 female) and one of their parents, primarily their mothers, were used for this data analysis.
Measures
Memorial Symptom Assessment Scale (MSAS 10–18)The MSAS 10–181 was modified from an adult measure, the MSAS,13 for pediatric oncology patient use. The procedure for the translation of MSAS 10–18 to the Chinese version is described in a previous study.14 The MSAS was designed to capture symptoms with two parts: the first part was used to assess the occurrence of a symptom and the second part was used to assess three dimensions of each symptom, including frequency, severity, and distress.13 If a symptom was reported, then a separate 4-point Likert scale (for frequency and severity) and a 5-point Likert scale (for distress) were used. Twenty-four items were evaluated for three dimensions, and eight items were evaluated for severity and distress only because frequency was not relevant (such as hair loss).13 A total score for each symptom was computed for each dimension (i.e., frequency, severity, and distress). The MSAS 10–18 includes three subscales: 1) Psychological subscale (PSYCH) (six items of psychological symptoms, including feeling sad, worry, feeling irritable, feeling nervous, difficulty sleeping, and difficulty concentrating), 2) Physical subscale (PHYS) (11 items of physical symptoms, including lack of appetite, lack of energy, pain, feeling drowsy, constipation, dry mouth, nausea, vomiting, change in taste, weight loss, and dizziness), and 3) the Global Distress Index (10 very frequently reported symptoms).13 The total of the MSAS score is computed as the average of the symptom scores for all the items.
Quality of Life for Children with Cancer (QOLCC)The QOLCC, a moderate-length scale, consists of 34 items used for examining both generic (physical, psychological, cognitive, and social functioning) and disease-specific domains (disease/symptom).15, 16, 17 It also assesses the understanding of illness and communication between parents and health care professionals. Moreover, it takes into consideration the developmental ability of children to respond and, therefore, includes three versions with identical content, one for children (who are younger than 12 years), one for adolescents (who are 12 years or older), and one for parents. For each item, patients were presented with four graded responses: 0
=never, 1=sometimes, 2=often, or 3=always being a problem. The sum of items in each subscale gives a score for each domain, including physical (five items), psychological (six items), cognitive (six items), social (four items), and disease/treatment (four items). For the purpose of this study, the subscales of understanding (four items) and communication (five items) were not used. Higher scores reflect more symptoms or health-related problems. Hence, the higher the score, the more negative the impact on the patient and thus a lower quality of life. The reliability of each subscale ranged from 0.69 to 0.79.16 The construct validity was examined by confirmatory factor analysis, and all of the subscales reached the criteria for the goodness of fit model (including nonsignificant χ2, χ2/df less than 3, acceptable for the goodness of fit index (GFI), the adjusted GFI, and the root mean square error of approximation).15 The feasibility of the QOLCC was demonstrated by a high rate of data completeness, acceptable floor and ceiling effects, and clinical validity for on- and off-treatment groups.17
The PedsQL Multidimensional Fatigue Scale was used to measure patients' self-reported fatigue symptoms.18 It consists of three subscales totaling 18 items. They include “general fatigue” (six items), “sleep/rest fatigue” (six items), and “cognitive fatigue” (six items). All items used a 5-point Likert response set, ranging from 1=“not at all” to 5=“always.” The development of the PedsQL Multidimensional Fatigue Scale included an intensive literature review and focus group discussions.18 Two versions of the same Fatigue Scale were used, one for young children and the other for adolescents. The version for young children was simplified linguistically in consideration of the cognitive ability of this age group. The instrument was then pilot-tested with three children with cancer. In the original instrument, higher scores indicated fewer problems with or symptoms of fatigue. To assess the fatigue intensity for patients, the PedsQL Multidimensional Fatigue Scale used in the current analysis was rescored so that higher scores indicated that more severe fatigue was perceived by patients. Cronbach's alpha for each subscale ranged from 0.78 to 0.95 for this study.
The CBCL is a widely accepted, well-validated, 113-item, child-completed behavior rating scale (age 6–18 years).19 The CBCL/6–18 is modified from the CBCL/4–18.20 The CBCL/6–18 has 118 problem items that contribute to the social competence index, eight narrow-band syndromes (“anxious/depressed,” “withdrawn/depressed,” “somatic complaints,” “social problems,” “thought problems,” “attention problems,” “rule-breaking behaviors,” and “aggressive behavior”), and other problems. It assesses the “internalizing” and “externalizing” behavior problems and social competencies indexed. Each item was scored “0” if the problem was “not true” of the child, “1” if the problem was “somewhat” or “sometimes true,” or “2” if it was “very true” or “often true.” The total score of each narrow-band and broad-band syndrome was computed for each subject. Studies have shown good reliability and validity with the Chinese version of CBCL/4–18 in Taiwanese children.21, 22 For the purpose of the present study, only the subscales, “anxious/depressed” (13 items) and “withdrawn/depressed” (eight items), were used as “internalizing behaviors.”
Demographic InformationDemographic characteristics were collected from medical records, including age, gender, disease information, and treatment information. Demographic data were also collected from parents.
Procedures
The study was reviewed by the Institutional Review Board of Chang Gung Children's Hospital before the study. The hospital's established procedures for protecting confidentiality were strictly followed. Detailed procedures for the recruitment process in this study were described in another study.14
Data Analysis
Descriptive statistics were calculated for the demographic characteristics of patients and their mothers. Frequency, severity, and distress of each symptom also were collected by the instruments described above. In addition, their disease status and type of cancer were collected from medical records. The internal consistency of each subscale and the total scales was examined using Cronbach's alpha.23 To evaluate the independence and importance of the three dimensions (frequency, severity, and distress) assessed by the MSAS 10–18, three statistical methods were used. First, mean scores of each dimension (frequency, severity, and distress) were computed and compared. Second, simple correlation coefficient (Spearman) was used to examine the overall correlations with each other. Canonical correlation was used to examine the relationships among and between the subscales of frequency, severity, and distress for the PHYS and the PSYCH of MSAS 10–18 with quality of life, fatigue, and internalizing behaviors. In addition to canonical correlation, partial canonical correlation also was used to determine which subscale of frequency, severity, or distress provides more information for symptom assessment. To identify the occurrence of symptoms experienced by Taiwanese pediatric oncology patients, our data also were compared with those of another published study that used the same instrument, the Collins et al.1 study data from the United States. Student's t-test using the Bonferroni correction was calculated to compare the occurrence of each symptom.
Results
Symptom Occurrence, Characteristics, and Distress by Treatment Status and Cancer Type
Detailed information regarding symptom prevalence by treatment status and type of cancer is presented in Table 1. The mean (standard deviation [SD]) number of symptoms reported per patient in the on-treatment group was 10.01 (6.24), which was significantly greater than that reported in the “off-treatment group” (mean=7.50, SD=6.69) (P=0.04). In general, patients in the on-treatment group reported statistically more distressing symptoms, including “nausea,” “lack of appetite,” “difficulty swallowing,” “change in how food tastes,” “weight loss,” and “hair loss,” than those in the off-treatment group. However, patients in the off-treatment group reported more distressing symptoms in “lack of concentration,” “feeling nervous,” “feeling drowsy,” “feeling sad,” and “feeling irritable.” Table 1 also lists the occurrence of individual symptoms by type of cancer. There were no statistical differences among distressing symptoms among the three types of cancer. Table 2 lists the occurrence and characteristics of symptoms reported by pediatric oncology patients. Symptoms that occur with higher frequency did not necessarily rank as high in severity or distress (one example is “worry”).
Table 1. Occurrence of Symptoms for Pediatric Oncology Patients by Treatment Status and Cancer Diagnosis
| Symptom | On-Treatment (n=108) | Off-Treatment (n=36) | Leukemia (n=75) | Lymphoma (n=26) | Solid Tumor (n=43) |
|---|---|---|---|---|---|
| n (%) | n (%) | n (%) | n (%) | n (%) | |
| Lack of concentration | 42 (38.9) | 17 (47.2) | 36 (48.0) | 5 (19.2) | 18 (41.9) |
| Pain | 49 (45.4) | 10 (27.8) | 29 (38.7) | 9 (34.6) | 21 (48.8) |
| Lack of energy | 57 (52.8) | 18 (50.0) | 43 (57.3) | 12 (46.2) | 20 (46.5) |
| Cough | 36 (33.3) | 9 (25.0) | 23 (30.7) | 8 (30.8) | 14 (32.6) |
| Feeling nervous | 11 (10.2) | 4 (11.1) | 12 (16.0) | 0 (0.0) | 3 (7.0) |
| Dry mouth | 49 (45.4) | 11 (30.6) | 33 (44.0) | 10 (38.5) | 17 (39.5) |
| Nauseaa | 55 (50.9) | 6 (16.7) | 33 (44.0) | 7 (26.9) | 21 (48.8) |
| Feeling drowsy | 50 (46.3) | 18 (50.0) | 41 (54.7) | 9 (34.6) | 18 (41.9) |
| Numbness/tingling | 33 (30.6) | 6 (16.7) | 21 (28.0) | 6 (23.1) | 12 (27.9) |
| Insomnia | 36 (33.3) | 9 (25.0) | 27 (36.0) | 5 (19.2) | 13 (30.2) |
| Problems with urination | 13 (12.0) | 3 (8.3) | 9 (12.0) | 1 (3.8) | 6 (14.0) |
| Dyspnea | 18 (16.7) | 3 (8.3) | 15 (20.0) | 2 (7.7) | 4 (9.3) |
| Diarrhea | 37 (34.3) | 12 (33.3) | 29 (38.7) | 9 (34.6) | 11 (25.6) |
| Feeling sad | 20 (18.5) | 11 (30.6) | 20 (26.7) | 5 (19.2) | 6 (14.0) |
| Sweating | 50 (46.3) | 16 (44.4) | 36 (48.0) | 12 (46.2) | 18 (41.9) |
| Worrying | 50 (46.3) | 16 (44.4) | 38 (50.7) | 12 (46.2) | 16 (37.2) |
| Itching | 36 (33.3) | 10 (27.8) | 22 (29.3) | 8 (30.8) | 16 (37.2) |
| Lack of appetiteb | 59 (54.6) | 13 (36.1) | 43 (57.3) | 9 (34.6) | 20 (46.5) |
| Dizziness | 32 (29.6) | 8 (22.2) | 25 (33.3) | 5 (19.2) | 10 (23.3) |
| Difficulty swallowingb | 16 (14.8) | 1 (2.8) | 9 (12.0) | 4 (15.4) | 4 (9.3) |
| Feeling irritable | 28 (25.9) | 15 (41.7) | 30 (40.0) | 5 (19.2) | 8 (18.6) |
| Vomitingc | 45 (41.7) | 6 (16.7) | 26 (34.7) | 7 (26.9) | 18 (41.9) |
| Mouth sores | 21 (19.4) | 4 (11.1) | 15 (20.0) | 3 (11.5) | 7 (16.3) |
| Change in the way food tastesb | 35 (32.4) | 5 (13.9) | 27 (36.0) | 5 (19.2) | 8 (18.6) |
| Weight lossb | 45 (41.7) | 8 (22.2) | 29 (38.7) | 8 (30.8) | 16 (37.2) |
| Hair lossb | 44 (40.7) | 7 (19.4) | 31 (41.3) | 5 (19.2) | 15 (34.9) |
| Constipation | 29 (26.9) | 7 (19.4) | 23 (30.7) | 5 (19.2) | 8 (18.6) |
| Swelling of arms/legs | 19 (17.6) | 4 (11.1) | 13 (17.3) | 5 (19.2) | 5 (11.6) |
| “Don't look like myself” | 33 (30.6) | 7 (19.4) | 27 (36.0) | 4 (15.4) | 9 (20.9) |
| Skin changes | 33 (30.6) | 6 (16.7) | 24 (32.0) | 8 (30.8) | 7 (16.3) |
aP |
bP |
cP |
Table 2. Occurrence and Characteristics of Symptoms in Pediatric Oncology Patients
| Symptom | Overall Occurrence (%) | Frequencya (%) | Severityb (%) | Distressc (%) |
|---|---|---|---|---|
| Lack of concentration | 41.0 | 25.9 | 7.1 | 3.4 |
| Pain | 41.0 | 34.9 | 16.8 | 19.0 |
| Lack of energy | 52.1 | 50.0 | 20.5 | 14.7 |
| Cough | 31.3 | 22.0 | 7.1 | 8.9 |
| Feeling nervous | 10.4 | 6.3 | 1.8 | 0.0 |
| Dry mouth | 41.7 | 36.8 | 8.8 | 1.7 |
| Nausea | 42.4 | 35.0 | 19.3 | 23.0 |
| Feeling drowsy | 47.2 | 44.0 | 21.2 | 7.4 |
| Numbness/tingling in hands/feet | 27.1 | 18.9 | 9.0 | 5.3 |
| Insomnia | 31.3 | 22.1 | 15.0 | 15.6 |
| Problems with urination | 11.1 | 7.3 | 4.4 | 6.3 |
| Dyspnea | 14.6 | 10.7 | 3.5 | 4.8 |
| Diarrhea | 34 | 29.4 | 12.4 | 6.1 |
| Feeling sad | 21.5 | 16.4 | 4.4 | 9.7 |
| Sweating | 45.8 | 37.9 | 13.4 | 4.6 |
| Worrying | 45.8 | 38.2 | 15.0 | 1.5 |
| Itching | 31.9 | 29.4 | 7.1 | 9.9 |
| Lack of appetite | 50.0 | 48.5 | 20.5 | 15.3 |
| Dizziness | 27.8 | 19.4 | 9.8 | 7.5 |
| Difficulty swallowing | 11.8 | 8.2 | 2.7 | 31.3 |
| Feeling irritable | 29.9 | 19.8 | 5.3 | 9.3 |
| Vomiting | 35.4 | 25.7 | 16.8 | 18.0 |
| Mouth sores | 17.4 | 4.4 | 28.0 | |
| Change in the way food tastes | 27.8 | 12.5 | 12.5 | |
| Weight loss | 36.8 | 8.0 | 11.3 | |
| Hair loss | 35.4 | 10.7 | 29.4 | |
| Constipation | 25.0 | 8.8 | 13.9 | |
| Swelling of arms/legs | 16.0 | 4.4 | 13.0 | |
| “I don't look like myself” | 27.8 | 13.3 | 20.0 | |
| Skin changes | 27.1 | 5.4 | 7.7 |
aPercentage of rating from “a lot” to “almost always.” |
bPercentage of rating from “moderate” to “very severe.” |
cPercentage of rating from “quite a bit” to “very much.” |
Dimensionality of the MSAS 10–18
To determine the dimensionality of the MSAS 10–18, the mean scores of each dimension (including frequency, severity, and distress) were computed only for those symptoms actually experienced by the patients. For all patients, the correlation between mean severity scores and mean frequency scores across symptoms was r=0.96, the correlation between mean severity scores and mean distress scores was r=0.96, and the correlation between mean frequency scores and mean distress scores was r=0.95. These correlations suggest that frequency, severity, and distress were highly correlated to each other. Spearman’s correlation coefficient was used to examine the pairwise correlations among the three dimensions. The average Spearman’s correlation coefficient between severity and frequency for the 22 symptoms graded according to these two dimensions was r=0.98 (range 0.95–1.00). The average correlation for severity and distress for 30 symptoms was r=0.94 (range 0.91–1.00), and the average correlation for frequency and distress was r=0.97 (range 0.92–1.00).
Moreover, separate canonical correlations were calculated for overall quality of life, fatigue and internalizing behaviors, scores for each severity, frequency and distress to decide whether each of the frequency, severity, and distress subscales could add more information to them. The scores for each of the symptoms, that is, PHYS and PSYCH, were calculated as frequency, severity, and distress scores. Separate canonical correlations were computed for each frequency, severity, and distress, when compared with another set of canonical variables (i.e., quality of life, fatigue, and internalizing behaviors). The canonical correlation coefficients were 0.51 for the set of frequency scores, 0.54 for severity scores, and 0.50 for distress scores (all Ps<0.05). These findings suggest that severity scores was the most informative single dimension.
Furthermore, we used partial canonical correlation to evaluate whether the frequency or distress subscale scores added information beyond that contained in the severity score. The canonical correlations between the severity score of the two MSAS 10–18 subscales and the scores on the quality of life, fatigue, and internalizing behaviors were computed after partialling out the two frequency and severity subscale scores. The findings show no significant canonical correlation when the severity subscale scores were partialled out from the comparison between the frequency subscale (P=0.06), but the distress subscale was significant (r=0.40, P<0.05). Combined with the original canonical correlation analysis, this finding suggests that the severity and distress measures, but not frequency measure, can provide the most information about quality of life, fatigue, and internalizing behaviors.
Table 3 lists the prevalence of symptoms experienced by older children in the present study as well as frequency data from another study.1 The Collins et al. study described the prevalence and characteristics of symptoms from 160 pediatric oncology patients aged 10–18 years in the United States. Taiwanese pediatric patients aged 10–18 years reported significantly higher rates of physical symptoms, including “lack of concentration,” “dry mouth,” “diarrhea,” “sweating,” “change in the way food tastes,” “constipation,” and “don't look like myself” (P<0.05). In contrast, Taiwanese adolescents reported a significantly lower occurrence of psychological symptoms, such as “feeling nervous” and “feeling sad” (P<0.05).
Table 3. Comparison of Symptom Prevalence (%) of the Present Study with the Study of Collins et al.1
| Symptom | Collins et al. (n=159) | Present Study (n=144) | t-Test |
|---|---|---|---|
| Lack of concentration | 20.1 | 41.0 | −3.96a |
| Pain | 49.1 | 41.0 | 1.41 |
| Lack of energy | 49.7 | 52.1 | −0.42 |
| Cough | 40.9 | 31.3 | 1.74 |
| Feeling nervous | 35.8 | 10.4 | 5.19a |
| Dry mouth | 30.8 | 41.7 | −1.97c |
| Nausea | 44.7 | 42.4 | 0.40 |
| Feeling drowsy | 48.4 | 47.2 | 0.21 |
| Numbness/tingling in hands/feet | 22.0 | 27.1 | −1.03 |
| Insomnia | 30.8 | 31.3 | −0.09 |
| Problems with urination | 6.3 | 11.1 | −1.49 |
| Dyspnea | 16.5 | 14.6 | 0.46 |
| Diarrhea | 20.1 | 34 | −2.73b |
| Feeling sad | 35.8 | 21.5 | 2.74b |
| Sweating | 20.3 | 45.8 | −4.74a |
| Worrying | 35.4 | 45.8 | −1.84 |
| Itching | 32.7 | 31.9 | 0.15 |
| Lack of appetite | 39.6 | 50.0 | −1.82 |
| Dizziness | 24.5 | 27.8 | −0.65 |
| Difficulty swallowing | 12.6 | 11.8 | 0.21 |
| Feeling irritable | 34.6 | 29.9 | 0.87 |
| Vomiting | 27.7 | 35.4 | −1.44 |
| Mouth sores | 13.9 | 17.4 | −0.84 |
| Change in the way food tastes | 16.5 | 27.8 | −2.38c |
| Weight loss | 26.6 | 36.8 | −1.91 |
| Hair loss | 28.3 | 35.4 | −1.33 |
| Constipation | 13.8 | 25.0 | −2.47c |
| Swelling of arms/legs | 12.0 | 16.0 | −1.01 |
| “I don't look like myself” | 15.8 | 27.8 | −2.54c |
| Skin changes | 20.1 | 27.1 | −1.44 |
aP |
bP |
cP |
Discussion
This is the first article to investigate the occurrence of symptoms experienced by 10- to 18-year-old Taiwanese cancer patients. Symptoms are defined as subjective constructs, and only the person who experiences it can be considered a valid reporter of the symptom.24 This study used the MSAS 10–18 to evaluate these subjects' symptom distress. The occurrence of symptoms for these Taiwanese cancer patients aged 10–18 years is different from that reported in Western studies. Different from the study of Collins et al.,1 the current study found that Taiwanese pediatric cancer patients, who are aged 10–18 years, reported a higher occurrence of two symptoms: “lack of energy “and “lack of appetite.” These are interesting study findings, and they can be explained by Chinese medicine philosophy. In Taiwanese culture, when a person is sick, the most common description of one's sickness is “there is no Yuan Qi (vitality).” Such a description of one's sickness has its origins in traditional Chinese medicine—there is an imbalance of “yin and yang” when someone is sick. When a person does not feel well or is sick, Chinese food therapy, a practice of healing using natural foods, is commonly used in Taiwanese culture. Together, it is not surprising that these two symptoms (lack of energy and lack of appetite) are the most prevalent distress for the older Taiwanese children who participated in this study. Although pain is also reported by this population, it is not emphasized as much as lack of energy and lack of appetite. Pain is recognized as an unavoidable symptom and should be endured in Taiwanese culture.2 Symptoms are experienced differently by each person, and the findings of this study confirm the importance of frequent assessment by health care professionals.
Researchers and clinicians continue to argue about the use of simple or multidimensional assessment tools for cancer patients.13, 25 Ideally, assessing symptoms having multiple dimensions (i.e., frequency, severity, and distress) can provide richer information than single-dimension assessments. High occurrence of a symptom does not necessarily mean more distress for patients and vice versa. For example, the item “difficulty swallowing” was only reported as modest in terms of frequency, but it was reported as the highest in terms of distress for patients in this study. However, assessing these dimensions using patient questionnaires may significantly increase the time burden for patients. This is particularly critical when patients are younger and have limited cognitive ability or are too sick to participate. During the data collection for this study, the subjects expressed their perception that the three subscales of frequency, severity, and distress look similar to them. They reported finding it difficult to distinguish the differences when they were filling out the questionnaire. This response from the subjects raises the issue of whether the assessment of the multiple dimensions of symptoms is necessary in the clinical setting.
Different from the Portenoy et al.13 study findings of distress subscale among frequency, severity, and distress subscales, this study found that the severity subscale is the most informative single dimension for MSAS 10–18. In addition to the severity subscale, knowing the level of distress can add more information to the accuracy of symptom assessment. Before interpreting these findings, we need to be aware of two issues: scoring of the scales and the format of the questionnaire. The scoring of symptoms using a scale of 1–4 or 1–5 may not be the best method to evaluate younger patients. In a comparison study of the Likert scale, the Visual Analog Scale (VAS), and the numeric VAS in 122 children from a pediatric outpatient clinic, the Likert scale was recommended for use in questionnaires for children.26 However, in a review study of comparisons among the three pain rating scales (i.e., VAS, the Verbal Rating Scale, and the Numerical Rating Scale), all the three are reliable and appropriate to use in clinical practice, but the Numerical Rating Scale has better sensitivity for statistical analysis.27 Therefore, the scoring system of a Numerical Rating Scale may be the best method to assess the symptoms for this population.
Second, the format of the MSAS 10–18 may have had an impact on the differentiation of the subscale of frequency, severity, and distress. In the MSAS 10–18, the frequency subscale “how often do you have it” is listed after the occurrence of a symptom (yes or no), followed by “how severe was it usually” and “how much did it distress or bother you,” respectively. Theoretically, an ideal symptom distress assessment tool should include the frequency (how often a symptom occurred), the severity, and level of distress. However, having such complicated concepts embedded in one symptom may not be appropriate for fieldwork with younger patients. If it is necessary that symptoms be evaluated using multiple dimensions or at least severity and distress, it is necessary to rearrange the format of the MSAS 10–18. One suggestion is that the MSAS items could be separated so that each subscale (i.e., frequency, severity, or distress) can be completed and evaluated separately.
In our study, completing the MSAS 10–18 required approximately 15minutes. However, in addition to symptom assessment, other self-reported measures, such as quality of life, fatigue, and behaviors for patients, may also be needed both in health outcome research and in clinical practices. Those measures, such as perceived stress or quality of life, are all subjective and need to be completed by the patients themselves. If the MSAS 10–18 is changed, the scoring system would be different, and new studies will be needed to evaluate its reliability and validity. The present study provides findings regarding the methodological issues surrounding symptom measurement, and together with these other studies suggest the value of assessing the multiple dimensions of symptoms.1, 13
Acknowledgment
Special thanks to Dr. Susan Jay for editing the manuscript.
References
- The measurement of symptoms in children with cancer. J Pain Symptom Manage. 2000;19:363–377
- . Adaptation in children with cancer: research with Roy's model. Nurs Sci Q. 2001;14:141–148
- . Life experience of Taiwanese adolescents with cancer. Scand J Caring Sci. 2002;16:232–239
- . Symptoms in children/young people with progressive malignant disease: United Kingdom Children's Cancer Study Group/Paediatric Oncology Nurses Forum Survey. Pediatrics. 2006;117:e1179–e1186
- . Symptoms affecting children with malignancies during the last month of life: a nationwide follow-up. Pediatrics. 2006;117:1314–1320
- . Managing cancer side effects to improve quality of life: a cancer psychoeducation program. Cancer Nurs. 2003;26:37–46
- . Health-related quality of life in pediatric patients with cancer. A structural equation approach with the Roy adaptation model. Cancer Nurs. 2002;25:74–80
- . Symptom monitoring and dependent care during cancer treatment in children: pilot study. Cancer Nurs. 2006;29:188–197
- . New measure of daily coping: development and preliminary results. J Pers Soc Psychol. 1984;46:892–906
- Symptoms in the palliative phase of children with cancer. Pediatr Blood Cancer. 2007;49:160–165
- . Distressing events for children and adolescents with cancer: child, parent, and nurse perceptions. J Pediatr Oncol Nurs. 2003;20:120–132
- . Symptoms of HIV disease. Palliat Med. 1991;5:46–51
- The Memorial Symptom Assessment Scale: an instrument for the evaluation of symptom prevalence, characteristics and distress. Eur J Cancer. 1994;30A:1326–1336
- . Symptom clustering in older Taiwanese children with cancer. Oncol Nurs Forum. 2008;35:273–281
- . Construct validity of newly developed quality of life assessment instrument for child and adolescent cancer patients in Taiwan. Psychooncology. 2003;12:345–356
- . The quality of life for cancer children (QOLCC) in Taiwan (part I): reliability and construct validity by confirmatory factor analysis. Psychooncology. 2004;13:161–170
- . The quality of life for cancer children (QOLCC) for Taiwanese children with cancer (part II): feasibility, cross-informants variance and clinical validity. Psychooncology. 2004;13:171–176
- . The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality of Life Inventory Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module. Cancer. 2002;94:2090–2106
- . Manual for the ASEBA school-age forms & profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families; 2001;
- . Manual for the Child Behavior Checklist/4–18 and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry; 2001;
- . Comparison of emotional/behavioral problems between children with cancer and healthy controls. J Clin Nurs. 2008;17:304–311
- . Competence of and emotional/behavioral problems in pediatric oncology patients in Taiwan. Cancer Nurs. 2004;27:413–422
- . Coefficient alpha and the internal structure of tests. Psychometrika. 1951;16:297–334
- Advancing the science of symptom management. J Adv Nurs. 2001;33:668–676
- . Symptom clusters: establishing the link between clinical practice and symptom management research. Support Care Cancer. 2006;14:792–794
- . A comparison of Likert scale and visual analogue scales as response options in children's questionnaires. Acta Paediatr. 2004;93:830–835
- . Pain: a review of three commonly used pain rating scales. J Clin Nurs. 2005;14:798–804
This research was supported by a grant to Dr. Yeh from the National Health Research Institutes, Taiwan (grant no. NHRI-EX95-9302PI) and the National Science Council (NSC grant no. 94-2314-B-182-014).
PII: S0885-3924(09)00702-7
doi:10.1016/j.jpainsymman.2009.04.023
Crown Copyright © 2009. Published by Elsevier Inc. All rights reserved.
Volume 38, Issue 5 , Pages 738-746, November 2009
