«The Experience of a Sore Mouth and Associated Symptoms in Patients With Cancer Receiving Outpatient Chemotherapy By: Brown, Carlton G. PhD, RN, AOCN; ...»
In a large study (n = 599) of patients receiving chemotherapy for solid tumors, Elting and Shih 21 found that patients with OM were more likely to experience fatigue than those patients without OM (9% vs 5%; P = 0.007). Dodd et al 16 confirmed that outpatients in their study who developed OM had a significant increase in mood disturbances as compared with outpatients who did not have OM (P = 0.03) and also experienced significantly higher levels of depression (F1,75 = 13.47, P.001) and anger (F1,75 = 9.47, P = 0.02). In summary, there is initial empirical support for relationships among OM and pain, anxiety, depression, and fatigue in several samples of patients with cancer but little evidence for an association between OM and trouble sleeping.
There seem to be no other known studies that investigate the relationship between SM and various symptoms.
CONSEQUENCES OF OM
Patients have reported their experiences with OM to researchers, including consequences.3,4,38 For instance, patients who received radiotherapy (n = 33) for head and neck cancer identified OM as the most troublesome and debilitating of all adverse effects they experienced.3 Consequences of OM in these patients included difficulty eating/drinking (88%), weight loss (83%), depression (38%), difficulty talking (29%), sleep disturbances (25%), and hospitalization (13%).3 Among patients with weight loss, 29% had gastric tube placement to assist with oral intake. In this foundational research on OM, there is no evidence that any researchers have focused on the consequences of patient-reported SM.
Armstrong's SEM 5 served as the conceptual framework for the study. The term symptoms experience is conceptually defined using Armstrong's definition 5 for this study as "the perception of the frequency, intensity, distress, and meaning occurring as symptoms are produced and expressed."5 For this study, symptom experience is operationally defined in 2 ways: (1) severity and distress of individual symptoms (SM and 4 others-fatigue, feeling down/blue, feeling anxious, and trouble sleeping) and (2) a statistically significant association between any 2 symptoms.
Armstrong's SEM includes the overall symptoms experience, along with antecedents, defining attributes, and consequences of the symptoms and consequences of the overall symptoms experience. Armstrong used symptoms in the plural when discussing an experience. For the purposes of this study, symptom is used in the singular form because all the symptoms combined together are viewed as comprising a symptom experience.
Antecedents that affect the symptom experience fall into 3 major categories: demographic characteristics (age, sex, race), disease characteristics (type and state of disease, treatment, comorbidities), and individual characteristics (health knowledge, values, past experiences).5 In this study, antecedents included age, sex, marital status, type of cancer, and type of chemotherapy. Defining attributes include a symptom's frequency, intensity, distress, and meaning.5 In this study, defining attributes of symptoms were operationally defined as symptom occurrence, severity, and distress, since these were captured in the parent study dataset on a daily basis. Consequences, or negative overall effects of the symptom experience, include specific health outcomes in numerous populations, including those with cancer.5 Examples of consequences within the SEM include adjustment to illness, quality of life, mood, functional status, disease progression, and survival. In this study, the major consequence was impairment in functional status as measured by amount of oral liquid intake, time spent lying down, ability to work, and interference of symptoms with daily activity.
Following the recommendation of Magee et al 39 that a guiding conceptual model will promote a rigorous secondary data analysis and guard against threats to study validity (including external validity [ie, generalizability of findings]), the researchers used SEM 5 as the conceptual framework for analyzing the data for this respective study. Armstrong's SEM also helped organize and classify variables that potentially contributed to (antecedents) SM and its consequences (negative overall effects). Finally, SEM allowed for simultaneous investigation of the effects of severity and distress of SM and other symptoms on the overall SM experience.
Methods Design This study was a retrospective descriptive, correlational secondary analysis. Data were used from a prospectively conducted randomized clinical trial entitled "Telephone-Linked Care for Cancer Symptom Monitoring."40 The purpose of the parent study was to assess the efficacy of the telephone-linked care (TLC) chemotherapy alerting system in the symptom management of adults receiving chemotherapy for cancer. The TLC chemotherapy alerting system was a computer-based communication system, using a digitized human voice that recorded and monitored patients' cancer symptom experience through daily, automated telephone "conversations."
Mooney and colleagues 40 described in detail how the TLC script and interview guide were created and tested in pilot work. To assess patients' symptoms, a rating system of 1 to 10 was used with scores increasing as severity increased. An expert panel provided input into the symptom measurement system and drill-down questions if there were any symptoms reported. The script was pilot tested with 27 patients with cancer who reported a high satisfaction with TLC. The technique, TLC voice, and the duration of the calls were acceptable to patients without any difficulty understanding and responding to respective questions.
In consenting to the parent study, participants provided blanket consent for future secondary analyses such as this study. In this secondary analysis, all identifying information (name, social security number, address, etc) was separated from the data. The study was approved by the University of Utah Institutional Review Board.
Methods of the TLC parent study, along with specific elements that pertain to this secondary analysis, are discussed below.
Sample and Setting
The original TLC study sample consisted of 223 participants in 4 outpatient chemotherapy treatment centers across the United States who reported at least one of several targeted symptoms during the first cycle of chemotherapy. In addition, TLC patients needed to have daily access to a touch-tone telephone, understand and speak English, be 18 years or older, have no physical or mental conditions that would have prevented them from participating, and be beginning cycle 2 of cancer chemotherapy with or without biotherapy. Since the 223 participants were screened on all symptoms in the parent study, all of them were used for this secondary analysis, thus allowing an examination of symptom associations in patients who experienced SM and at least one other symptom of interest.
Variables and Measures
In the original TLC parent study,40 the researchers selected common chemotherapy-related symptoms and presented them to participants in everyday language. For example, instead of asking patients whether they were experiencing depression, anxiety, and/or OM, they used the terms feeling down /blue, feeling anxious, having trouble sleeping, and sore mouth. The TLC parent study investigators assumed that these descriptors represented depression, anxiety, insomnia, OM, respectively.
Participants were instructed to call one time per day during cycles 2 and 3 of chemotherapy. They were asked whether they had nausea and vomiting, had trouble sleeping, were experiencing fatigue, were feeling down/blue, were feeling anxious, had SM, had fever and chills, had diarrhea, had constipation, had pain, and had distress about appearance changes. If they answered in the affirmative for any symptom, they were then asked to rate its severity and distress separately using a scale of 1 (minimal) to 10 (worst). Later during the telephone call, the TLC system asked questions about specific consequences of the symptoms in 4 areas. First, patients reported how much their symptoms prevented them from doing their daily activities (on a scale of 0-10, with 0 meaning that the symptoms did not in any way prevent them from doing daily activities and 10 meaning that the symptoms completely prevented them from doing daily activities). Second, they reported the amount of time in hours they had spent lying down in the previous 24 hours. Third, patients who were actively working indicated whether they had worked the last day they were supposed to work. Fourth, they also reported the number of 8-oz glasses of liquid they consumed in 24 hours. In addition to symptom-related data, demographic and clinical data were collected from each patient, specifically age, sex, marital status, ethnical background, educational level, type of cancer, and type of chemotherapy. The number of days in a cycle of chemotherapy ranged from 21 to 28 days.
SPSS version 11.0 (SPSS Inc, Chicago, Illinois) was used to manage and analyze the data. To create measures of severity and distress across multiple daily calls, aggregated scores were computed similar to the method used by Cleeland and Ryan 41 in scoring the Brief Pain Inventory. In this method, the highest, average, and lowest (HAL) ratings of SM severity and distress across all of a given patient's calls were summed and averaged to obtain an SM severity and distress score forboth cycles of chemotherapy. This method captured the variability in individual ratings of all respective symptoms. For the consequence measures (interference of symptoms with daily activities, time spent lying down, and liquid intake), the same method was used to create a HAL score for daily activity, time spent lying down, and liquid intake.
If a patient called into the system and did not report SM, he or she was categorized as a nonresponder in terms of SM, even if at least 1 of the other 9 symptoms was reported. For instance, a patient could report nausea and vomiting and trouble sleeping but no SM. On days when patients did not call into the system at all (ie, missed a call), they were coded as missing. Descriptive statistics, independent t tests, and [chi]2 tests were used to analyze patients' characteristics and SM severity and distress. Analyses of relationships among SM and other variables were conducted using Pearson correlations, tests of differences in correlations, and paired t tests to examine differences in symptom severity across time.
Findings Specific Aim 1 Specific aim 1 evaluated the demographic and clinical factors associated with the occurrence of an SM. Of the 223 outpatients in the TLC study, 115 (51%) reported an SM at least once over 2 cycles (cycles 2 and 3) of chemotherapy. Table 1 presents the demographic and clinical characteristics of those patients who did (n = 115) and did not (n = 108) report an SM.
Table 1 Demographic and Clinical Characteristics of the Sample and Subgroups With and Without a Sore Mouth and Differences Between Groups ([chi]2) Those reporting an SM were younger (53.92 years) than those who did not (57.75 years) (t = 2.33, P = 0.21).
Participants with an SM were mostly women (79.5%), were white (94.4%), were married (75.5%), and had some college education (31%). When compared with patients without an SM, those reporting SM were more likely to be diagnosed with breast cancer or Non-Hodgkin's Lymphoma (NHL) ([chi]2 = 40.6, P =.001) and to receive doxorubicin and cyclophosphamide ([chi]2 = 9.17, P =.01) (Table 1).
Table 2 presents the HAL severity scores for SM, trouble sleeping, feeling anxious, feeling down/blue, and fatigue. The mean (SD) HAL SM severity score for cycle 2 was 3.1 (1.50), whereas for cycle 3, it was 3.09 (1.52), and these scores were not significantly different (t169 =.38, P =.90). Correlations between HAL SM severity score in cycle 2 and HAL SM severity score in cycle 3 revealed that scores were moderately and significantly correlated (r =.56, P.001).
Table 2 Mean and SD of Highest, Average, and Lowest (HAL) Severity and Distress Scores for Each Symptom (Possible Range = 1-10) Specific Aim 2 Specific aim 2 examined relationships among SM severity, SM distress, and the severity and distress of feeling blue, feeling anxious, trouble sleeping, and fatigue over cycles 2 and 3 of chemotherapy (see Table 3). The entire sample of patients (n = 223) were used in these correlations because this allowed examination of associations that included those who may or may not have experienced any of the symptoms of interest.
Statistically significant bivariate Pearson correlations (P.001 with a medium effect size) between SM severity and severity of other symptoms included trouble sleeping, r223 = 0.42; fatigue, r223 = 0.400; feeling down/blue, r223 = 0.36; and feeling anxious, r223 = 0.28. Statistically significant (P.001 with medium or low effect size) bivariate Pearson correlations between SM distress and distress of other symptoms included trouble sleeping, r223 = 0.23; fatigue, r223 = 0.27; feeling down/blue, r223 = 0.16; and feeling anxious, r223 = 0.15.
Table 3 Correlations of Sore Mouth (SM) Severity and Distress With Severity and Distress of Other Symptoms
Specific Aim 3