Research Article, J Food Nutr Disor Vol: 4 Issue: 2
Developing Theory-Based Measurement Tools for Improving Diet Compliance in Head and Neck Cancer Patients
Rogers LQ1*, Verhulst S2, Rao K3, Malone J4, Robbs R5 and Robbins KT6 | |
1Departmentf Nutrition Sciences, University of Alabama at Birmingham, 1720 2nd Avenue North, Webb 222, Birmingham, AL 35294-3360, USA | |
2Statistics and Informatics Core, Center for Clinical Research, Southern Illinois University School of Medicine, Springfield, IL, P.O. Box 19664, Springfield, IL 62794-9664, USA | |
3Department of Medicine and Simmons Cancer Institute, Southern Illinois University School of Medicine, P.O. Box 19677, Springfield, IL 62794-9677, USA | |
4Associated Otolaryngologists of Pennsylvania, Inc., 34 Northeast Drive, Hershey, PA 17033, USA | |
5Statistics and Informatics Core, Center for Clinical Research, Southern Illinois University School of Medicine, Springfield, IL, P.O. Box 19664, Springfield, IL 62794-9664, USA | |
6Department of Surgery and Simmons Cancer Institute, Southern Illinois University School of Medicine, P.O. Box 19677, Springfield, IL 62794-9677, USA | |
Corresponding author : Laura Q. Rogers University of Alabama Birmingham, Department of Nutrition Sciences, 1720 2nd Avenue North, Webb 222, Birmingham, AL 35294-3360,USA Tel: (205) 975-1667; Fax: (205) 934-7049 Email: rogersl@uab.edu |
|
Received: February 02, 2015 Accepted: March 30, 2015 Published: April 03, 2015 | |
Citation: Rogers LQ, Verhulst S, Rao K, Malone J, Robbs R, et al. (2015) Developing Theory-Based Measurement Tools for Improving Diet Compliance in Head and Neck Cancer Patients. J Food Nutr Disor 4:2. doi:10.4172/2324-9323.1000166 |
Abstract
Developing Theory-Based Measurement Tools for Improving Diet Compliance in Head and Neck Cancer Patients
Increasing compliance with diet recommendations is critical for improving health, quality of life, and survival among head and neck cancer (HNCa) patients. The social cognitive theory is a potentially useful behavioral theory framework for improving compliance but measurement tools assessing theory constructs related to diet compliance in HNCa patients are needed. Therefore, our study aim was to pilot test social cognitive theory construct measures, including prevalence and preliminary associations with self-reported compliance with physician or dietitian diet recommendations.Cross-sectional survey completed by 33 HNCa patients followed in an out-patient academic otolaryngology clinic.T
Keywords: Nutrition; Oncology; Supportive Care; Adherence; Health Behavior
Keywords |
|
Nutrition; Oncology; Supportive Care; Adherence; Health Behavior | |
Introduction |
|
Head and neck cancer (HNCa) patients may suffer significant weight loss due to side effects such as anorexia, taste alterations, mucositis, pain, and difficulty swallowing [1-4]. Counseling by a dietitian can reduce weight loss, enhance quality of life, and improve clinical outcomes [5-8] but up to 57% of HNCa patients may be noncompliant with such counseling [9,10]. Compliance can be improved by designing interventions based on a behavioral change theory [11]. For example, the social cognitive theory [12] provides principles that can be applied to assist individuals in making healthy behavior change [13]. Self-efficacy, the most well-known construct of the social cognitive theory, is an individual’s confidence in their ability to engage in a specified behavior [14]. Several other important social cognitive theory constructs include goal setting, outcome expectations (e.g., expected effects of the behavior), perceived barriers interference (e.g., how often barriers interfere with the behavior), environment (e.g., social support), and emotional coping responses (e.g., enjoyment of the behavior). Social cognitive theory constructs interact with each other in response to behavior choices and resultant experiences [11,13]. Although self-efficacy has a direct effect on behavior [15], it may also be influenced by and/or change behavior through other constructs. For example, improved self-efficacy may increase outcome expectations and decrease barriers interference which in turn is associated with increased adherence to the desired behavior [16]. Also, greater social support and enjoyment of the behavior may increase self-efficacy which is associated with greater behavior compliance [17]. | |
Few published diet interventions have used a behavior theory framework for improving diet compliance in HNCa patients on or off treatment. An intervention using one face-to-face session, one telephone counseling session, and three mailings was based on the social learning theory (renamed the social cognitive theory in 1986) [11,14,18]. This intervention focused on early stage HNCa patients who were post-treatment and at risk for a second cancer [18]. Another intervention focused on early stage post-treatment HNCa patients was based on the transtheoretical model [19]. This intervention used one face-to-face counseling session, weekly telephone sessions, and mailed materials [20]. Both studies reported significant improvements in fruit and vegetable intake in the intervention group [18,20]. | |
Identifying social cognitive theory constructs associated with diet compliance can be used to improve diet behavior change in future nutritional interventions for HNCa patients. Although social cognitive theory correlates have been studied in HNCa patients for outcomes other than diet compliance (i.e., emotional distress [21] and exercise [22]), no prior study has examined predictors or correlates of diet compliance in HNCa patients using any behavior change theory. The single study reporting patient compliance with physician provided diet recommendations reported that anxiety was the only correlate of compliance in 30 post-treatment HNCa patients [9]. Although anxiety is not a specific social cognitive theory construct, it can be considered an aspect of the emotional coping response construct [11]. Moreover, identifying behavior change theory correlates is warranted given their mediating role in lifestyle compliance in survivors of other cancer types (e.g., self-efficacy mediated diet behavior among breast and prostate cancer survivors [23] and perceived barriers interference mediated exercise behavior among breast cancer survivors [24]). | |
For some HNCa patients, adhering to recommended nutritional supplements with or without feeding tube use is an important aspect of diet compliance. Nine of 23 HNCa patients receiving chemoradiotherapy did not use the tube effectively and experienced significant declines in weight, body mass index, and lean body mass compared with those using their feeding tubes [25]. Reasons for feeding tube noncompliance included nausea and vomiting (n=2), low educational level or lack of awareness of nutrition importance (n=5), and reluctance (n=2). Although not reported within a theoretical framework, these reasons suggest useful social cognitive theory constructs such as perceived barriers interference (e.g., nausea) and outcome expectations (e.g., awareness of important nutrition benefits). Also related, eating difficulties in HNCa patients may negatively impact social interactions suggesting the potential importance of environment (e.g., social support) [26]. | |
Changing diet behavior in response to dietary counseling is critical for improving health, quality of life, and survival among HNCa patients. Interventions based on behavior change theory frameworks are needed to improve compliance. However, the question of which theory constructs should be targeted remains unanswered due to the lack of information about the associations between constructs and behavior in HNCa patients. Furthermore, no study, to date, has published the development and testing of the measurement tools required to optimize future diet compliance interventions for HNCa patients using a behavior change theory. Therefore, our study aim was to develop social cognitive theory-based measurement tools related to diet compliance among HNCa patients. This included item reduction and preliminary testing of reliability and associations with selfreported diet compliance. | |
Methods |
|
This cross-sectional survey study took place in an academic, head and neck oncology clinic. Study inclusion criteria were: 1) history of HNCa, 2) currently followed in the clinic for HNCa, 3) age 18 to 89 years, and 4) able to read and complete the survey. Institutional review board approval was obtained. Because no identifying information was collected, a waiver of consent was approved. Potential participants were identified by a research coordinator present in the clinical area; the coordinator provided the individual with a survey and cover letter explaining the study and voluntary nature of participation. Surveys were completed in the clinic on the day of enrollment. A convenience sample was used because it was not feasible for the coordinator to approach all potentially eligible patients on busy clinic days. Although the survey asked about diet and exercise behavior, only the diet-related analyses are reported here. | |
Participants self-reported demographics, medical information, compliance with diet recommendations (i.e., entered days per week followed a physician or dietitian’s diet recommendations in the past six weeks), and compliance with supplement recommendations (i.e., entered number of supplements per day recommended and ingested in the past six weeks). Similarly, participants with a feeding tube at the time of survey completion also entered the number of supplements per day recommended and ingested through the feeding tube in the past six weeks. | |
Social cognitive theory construct measures previously used in other populations [22,27-31] were adapted for disease outcomes and treatment side effects experienced by HNCa patients [1-3]. For goal setting [27], a 3-point Likert type scale asked participants their plans for following physician or dietitian dietary advice for the past six weeks, next six weeks, and next three months (i.e., 1=plan to follow recommendations less, 2=about the same, or 3=more). Participants were also asked their diet goals for the past month, next three months, and next six months on a 6-point Likert type scale (i.e., 0=do not have a specific diet goal, 1=do not plan to follow recommendations, 2=plan to follow recommendations 1 to 2 days per week, 3=plan to follow recommendations 3 to 4 days per week, 4=plan to follow recommendations 5 to 6 days per week, 5=plan to follow recommendations 7 days per week). Responses for the past six weeks’ time period were summed for testing the association between goal setting and compliance. | |
The barriers self-efficacy scale (i.e., confidence in ability to overcome barriers to following physician or dietitian diet recommendations during the past six weeks) [22,28,29] included 14 common diet barriers experienced by HNCa patients [1-3]. Confidence was rated on a Likert scale from 0% to 100% (0%=not at all confident to 100% = extremely confident) with the mean used for the analyses. The perceived barriers interference scales [22] asked participants to rate on a 5-point Likert scale (1=never to 5=very often) the frequency with which barriers interfered with following diet recommendations during the past six weeks (25 items), supplement use (6 items) and feeding tube use (3 items). Each barriers interference scale was summed for the analyses. | |
For social support [30,31], the frequency with which friends (two items) or family (two items) encouraged or offered to help the participant stick with or follow diet recommendations during the past six weeks was assessed by summing the 5-point Likert scale responses (0=none to 4=very often). The outcome expectations scale [32] asked participants to rate their agreement on a 5-point Likert scale (1=strongly disagree to 5=strongly agree) with the statement that following diet recommendations would result in 15 potential benefits. For associations with diet compliance, responses were summed. Also, participants were asked a single item asking for their agreement (1 = strongly disagree to 5=strongly agree) with the following: “I do not feel that I would experience any important benefits from following the dietitian’s recommendations”. Diet enjoyment was measured with a single question (5-point Likert scale; 1=agree to 5=disagree) asking participants to rate their agreement with the following: “I enjoy eating what the physician or dietitian recommends” [22]. Responses were reversed for the analysis so that higher scores indicated greater enjoyment. | |
Descriptive statistics assessed prevalence of responses. Items with low frequency of occurrence or overlapping concepts were removed in an effort to shorten scales with ≥ 14 items to reduce participant burden in future studies. Internal reliability of the scales with multiple items (original and reduced items scales) was examined by calculating Cronbach's alphas (coefficient representing how well the individual items correlate with each other within the scale; values of ≥.9, ≥.8, and ≥.7 are considered excellent, good, and acceptable, respectively; values <.5 are usually interpreted as unacceptable [33]). Associations with following diet recommendations were tested by dichotomizing responses as following recommendations seven days/week versus less than seven days/week due to the small sample size and response distribution (i.e., 13 reported following recommendations seven days/ week with six doing so less than seven days/week). Associations were examined using chi-square or independent groups t-test. If available for the construct, the reduced items scales were used when testing associations. | |
Results |
|
The response rate for the full survey was 74% with this manuscript reporting data from a subgroup analysis of participants who selfreported receiving diet recommendations, receiving supplement recommendations, and/or having a feeding tube in the past six weeks (i.e., 33 of 137 HNCa patients). Limiting the time frame for receipt of diet recommendations to the past six weeks was done to focus the participant’s attention on the recent past. | |
Sample characteristics for the 33 participants are provided in Table 1. Of note, six participants with a feeding tube failed to self-report diet or supplement recommendations in the past six weeks but were included in the analyses because all patients with a feeding tube receive supplement recommendations as part of their clinical care. | |
![]() |
Table 1: Characteristics for head and neck cancer patients with feeding tube and/or receiving diet and/or supplement recommendations from a physician or dietitian in the past six weeks (n=33). |
To create the reduced items barriers self-efficacy scale, the most prevalent barriers reported as interfering with following diet recommendations were included (i.e., no appetite, sore mouth, nausea, difficulty swallowing, painful swallowing, low on money, feel full quickly, difficulty chewing, does not taste good, dry mouth, a lot of phlegm in throat, heartburn or acid reflux). The outcome expectations scale was reduced based on overlapping constructs (e.g., feel stronger with muscle strength, have more energy with feel less tired, etc.) and importance for HNCa patients (e.g., weight gain). As a result the following items were retained for the reduced outcome expectations scale: feel stronger, look better, gain weight, feel less tired, improve body shape, reduce risk of disease, and feel better. | |
The reduced items barriers interference scale for compliance with diet recommendations included barriers reported by >50% of participants [i.e., inability to chew solid food, food does not taste good, choking, lack of appetite, sore mouth, dry mouth, phlegm production in mouth, difficulty swallowing, painful swallowing, cannot taste food, do not like taste of recommended items, feel full too quickly, nausea, heartburn or acid reflux, spouse (or other) does shopping] and relevance to lower socioeconomic populations (i.e., recommendations are too expensive) for 16 total items. The supplement and feeding tube use barriers interference scales were not reduced (i.e., supplement use items=I do not like the taste of the supplements, supplements cause stomach upset, supplements are too expensive, I cannot get the supplements, supplements cause diarrhea, and I cannot travel with the supplements; feeding tube use items=using the feeding tube is embarrassing, no one has shown me how to use my feeding tube, and I do not know how to use the feeding tube). | |
Table 2 provides the descriptive statistics and Cronbach's alphas (i.e., internal reliability) for the social cognitive theory scales related to diet compliance. Of the nine Cronbach’s alphas reported, none were in the unacceptable range of <.5, two were between .7 and .8 (acceptable), two were between .8 and .9 (good), and three were ≥.9 (excellent) [33]. Only 3 (16%) felt that they would not experience any important benefits from following diet recommendations. Goal setting is not included in Table 2 because Cronbach’s alphas do not apply. | |
![]() |
Table 2: Prevalence and, when appropriate, internal reliability (i.e., Cronbach’s alpha) for barriers self-efficacy, perceived barriers interference, social support, outcome expectations, and enjoyment related to compliance with diet recommendations in head and neck cancer patients. |
With regard to prevalence data, the percentage of participants planning to adhere to diet recommendations more often was 21% for the past six weeks, 50% for the next six weeks, and 67% for the next three months. The percent planning to follow diet recommendations 7 days a week was 38% for the past six weeks, 43% for next six weeks, and 64% for next three months. Table 3 provides associations between social cognitive theory constructs and compliance with diet recommendations. | |
![]() |
Table 3: Prevalence and, when appropriate, internal reliability (i.e., Cronbach’s alpha) for barriers self-efficacy, perceived barriers interference, social support, outcome expectations, and enjoyment related to compliance with diet recommendations in head and neck cancer patients. |
Because associations did not differ between original and reduced items scales, associations using the reduced items are reported. The association for feeding tube barriers interference is not provided because only two participants indicated receiving recommendations and completed the feeding tube barriers interference scale. Table 4 provides a listing of the final survey items recommended along with scoring instructions. | |
![]() |
Table 4: Final survey items recommended for use in future studies. |
Discussion |
|
The majority of social cognitive theory measures tested in this report demonstrated acceptable to excellent internal reliability with goal setting, enjoyment, social support, and barriers self-efficacy suggesting promising associations with diet recommendation compliance. Our data supports the potential usefulness of the social cognitive theory as a theoretical framework for improving diet adherence in HNCa patients. Moreover, the measurement tools in this report warrant further study and may be useful in future research examining the role of social cognitive theory in diet compliance in this cancer group. We acknowledge that the barriers interference scale for diet supplement use and the reduced items barriers interference scale for general diet recommendations demonstrated Cronbach’s alphas between .5 and .6 (poor). The diverse nature of the items included in these scales may have contributed, in part, to their lower internal reliability (e.g., experiencing diarrhea as a barrier to supplement use would not necessarily be expected to be associated with inability to pay for supplements). Future study should refine and reevaluate these scales including larger sample sizes and factor analyses to identify scales and subscales with improved internal reliability. Importantly, the scales related to barriers self-efficacy, social support, and outcome expectations demonstrated promise as useful tools for assessing and improving diet compliance in HNCa patients. | |
It is noteworthy that our study is the first to develop social cognitive measurement tools and explore the potential role of the social cognitive theory in diet compliance among HNCa patients. Importantly, our preliminary study supports the need and legitimacy of further testing these measurement tools. Also, our data generates hypotheses related to potential social cognitive theory targets for future interventions aimed to improve diet compliance in this cancer group. Specifically, such interventions should include behavior change strategies addressing the constructs with ≥ large effect size associations with compliance in our study (i.e., goal setting, barriers self-efficacy, friend social support, and diet enjoyment). The expectation of benefit from compliance with diet recommendations was high suggesting that HNCa patients know why they should follow recommendations but may require behavioral interventions addressing social cognitive theory constructs other than outcome expectations in order to improve diet compliance. | |
Also related to future intervention design, the most frequent barriers to diet compliance in our study are consistent with HNCa clinical characteristics and treatments (e.g., anatomic alterations, cancer cachexia) [2,3,34,35]. Similar to our results, taste has been reported as a major barrier to supplement use [34] and embarrassment a potential barrier to feeding tube use [36]. Given the importance of compliance with supplements and feeding tube use, research is needed to determine best methods for health care professionals and dietitians to ask about these barriers, legitimize the patient concerns, and provide helpful solutions. | |
Further research is needed to determine why the percentage of participants intending to comply with diet recommendations increased in the future. Prospective studies are needed to determine if intention predicts improved compliance or is simply procrastination. Possible social cognitive theory based reasons for procrastination may include a feeling that barriers will interfere less in the future or greater confidence in their ability to overcome barriers in the future rather than the present. Given the very strong association between goal setting and diet compliance, future interventions should address and emphasize this construct. Further research that includes details regarding the specific diet recommendations received and diet-related goals set is warranted and may improve our ability to use goal setting as an important behavior change tool in this population. | |
This preliminary study’s small sample size precluded factor analysis and limited study power for detecting statistically significant associations between social cognitive theory constructs and diet compliance. Although disease-related characteristics (e.g., disease stage, treatment type) were obtained by self-report due to logistical and budgetary constraints, such an approach resulted in missing cancer stage data and lack of specific treatment details beyond the general categories of chemotherapy, radiation, and/or surgery alone. This combined with the small sample size prevented testing for potential moderation of our findings by disease characteristics. The primary reason for our small sample size was the low prevalence of patients self-reporting receipt of physician or dietitian diet recommendations in the past six weeks. This may have occurred due to the larger percentage participants being off treatment. Alternatively, patients may not perceive general, brief statements about diet from their physician as diet recommendations and may not be able to differentiate a dietitian from other clinic staff (e.g., nursing staff). Also related, third party payers often do not cover the cost of dietitian counseling which may reduce the frequency of dietitian counseling. The fact that only 27% felt they had recently received diet recommendations reinforces the value of interventions increasing patient access to more intensive nutritional counseling. | |
Continued examination of social cognitive theory construct measurement and potential contribution to diet compliance in HNCa patients in larger studies is warranted. Reduced items scales can be used to lessen participant burden, which is especially important given the often overwhelming nature of HNCa diagnosis and treatment. Test-retest reliability evaluation is needed but may be difficult in patients on treatment given changes in clinical barriers over time. Factor analysis may prove useful in identifying subcategories of barriers or outcome expectations with stronger predictive properties relative to diet compliance. Lastly, determining the moderating effects of treatment status (on/off treatment), disease stage, treatment type, body mass index, age, gender, education, emotional well-being, alcohol use, and tobacco use on reliability and associations with diet compliance would improve our ability to optimally use the social cognitive theory to improve diet compliance and achieve resultant benefits in HNCa patients. | |
Acknowledgments |
|
This project was supported by institutional support from the University of Alabama at Birmingham and Southern Illinois University (SIU) School of Medicine. In-kind support from the Center for Clinical Research at SIU was also provided. The authors do not have any relationships to disclose which would cause a conflict of interest. | |
References |
|
|