Abstract
OBJECTIVE:
To determine how communicative participation is affected in patients with oral and oropharyngeal head and neck cancers (HNC) pre-treatment, and whether communication function predicts HNC-specific quality of life (QOL) before treatment, beyond known demographic, medical, psychosocial, and swallowing predictors.
STUDY DESIGN:
Cross-sectional study.
SETTING:
Tertiary care academic medical center.
SUBJECTS and METHODS:
Eighty-seven patients with primary oral (40.2%) or oropharyngeal (59.8%) HNC were recruited prior to treatment. T stage, tumor site, and p16 status were extracted from medical records. Demographic and patient-reported measures were obtained. Communicative participation was measured using the Communicative Participation Item Bank (CPIB) General short form. A hierarchical regression analysis included demographic, medical, psychosocial, and functional measures of swallowing and communication as predictors; the University of Washington Quality of Life (UW-QOL v4) composite score was the predicted variable.
RESULTS:
Median baseline CPIB scores were 71.0 (SD = 11.83); patients with oral cancers reported worse scores. A final sequential hierarchical regression model that included all variables explained 71% of variance in QOL scores. Tumor site, T stage, and p16 status accounted for 28% of variance (P < 0.001). Perceived depression predicted an additional 28% of the variance (P < 0.001). Swallowing and communicative participation together predicted an additional 12% of variance (P = 0.005). Tumor site, perceived depression, swallowing, and communication measures were unique predictors in the final model. Finally, communicative participation uniquely predicted QOL, above and beyond other predictors.
CONCLUSION:
Pre-treatment communication predicted QOL, and was negatively impacted in some oral and oropharyngeal HNC patients.
Keywords: head and neck cancer, cancer outcomes, quality of life, communication disorders
Introduction
Patients with head and neck cancer (HNC) often experience significant deficits in verbal communication. Posttreatment speech problems are commonly higher in oral versus oropharyngeal cancers, with prevalence ranging from 30% to 60% in oropharyngeal and oral cancer patients respectively, depending upon how speech impairments are measured.1 Difficulties with communication may lead to isolation and negatively impact quality of life (QOL).2,3
While long-term communication deficits in patients with oral and oropharyngeal HNC are relatively well known,1,4–6 very few studies have investigated communication function at baseline or pre-treatment.2,7 Without pre-treatment measures, there is no way to parse out the effect of the tumor on communication function at baseline, independent from treatment effects. This makes it difficult to compare the efficacy of different treatment modalities on functional communication outcomes. Further, without baseline measures, there is no way to measure changes in communication over time.
One challenge to establishing baseline communication function has been the relative lack of validated patient-reported outcome measures (PROMs) dedicated towards measuring this construct.8 Many HNC-specific QOL PROMs only include one or two items dedicated to speech or voice. These PROMs may lack sensitivity to broader patient concerns about voice or speech, and are inadequate for making clinical and research recommendations.9,10 An additional limitation in HNC research has been the relative lack of PROMs that capture functional speech, as opposed to voice-related difficulties.2 While voice-related QOL measures may sensitively measure difficulties for those with laryngeal cancer, those with oral and oropharyngeal-based tumors may have more deficits in speech-related tasks due to surgical or radiation effects on structures in the oral and pharyngeal cavities.11
To address this gap, our team developed a PROM called the Communicative Participation Item Bank (CPIB).12 The CPIB measures the impact of HNC on communication in everyday settings, or “communicative participation”. Items in the CPIB were developed on the basis of a literature review8 and qualitative studies.13,14 The CPIB was subsequently validated in several large-scale psychometric studies across a number of clinical populations, including HNC,12,15,16 and has been used to investigate relationships between communicative participation and other HNC outcomes such as speech intelligibility,17 and depression.18,19 One advantage of the CPIB is that it is able to capture communication difficulties, regardless of whether deficits are related to speech, voice, or other contributing factors. The use of a single PROM that applies to multiple HNC populations obviates the need for clinicians to use multiple instruments and reduces overall patient burden.20
While communicative participation is an important concern of oral and oropharyngeal HNC patients,1,21 it is unknown how this factor contributes to HNC-specific QOL at pre-treatment. It also is unknown how this type of PROM might predict QOL at pre-treatment, above and beyond other known predictors, such as demographic factors (age, sex), medical factors (tumor site, T stage, p16 status), psychosocial factors such as depression, or functional measures of swallowing.22–26 Given that baseline measures of HNC-specific QOL have been identified as significant predictors of QOL following treatment for HNC, understanding the contributions to QOL at baseline is important for treatment planning and for counseling individual patients, as well as for designing and interpreting research.27 Thus, this study determined: 1) how communicative participation is affected in patients with oral and oropharyngeal HNC before treatment and whether it varies by tumor site; and 2) if communicative participation predicts HNC-specific QOL before treatment, above and beyond other known demographic and medical factors, perceived depression, and swallowing-related QOL.
Methods
Design
This was a prospective cross-sectional study of patients with oral and oropharyngeal cancer prior to treatment. Primary measures of interest were the CPIB General short form scores and the UW-QOL composite scores prior to treatment. QOL predictor variables were age, sex, p16 status, tumor site, T stage, perceived depression, and functional measures of swallowing and communication. This study was approved by the University of Washington Institutional Review Board.
Sample and Setting
Participants in this study included a subsample of patients with HNC enrolled in the University of Washington’s ongoing Communication Outcomes after Head and Neck Cancer study (NIH/NCI:R01CA177635 PI: Eadie). Patients were recruited from the University of Washington Medical Center, an academic tertiary care facility, between 2014 and 2019. To be included in the broader study, patients met the following criteria: a) age 18+ years; b) received no treatment for head and neck cancer in the past; c) use speech as their primary mode of communication; and d) able to speak and read English well enough to complete self-report questionnaires. Participants were excluded if they: a) had previously altered anatomy of the upper aerodigestive tract; b) had pre-existing speech or voice impairments unrelated to the tumor, (e.g., neurologic disorders such as Parkinson’s disease or stroke); c) were unable to give informed consent; or d) were unable to complete self-administered questionnaires written in simple English for cognitive, psychiatric, or other reasons.
Of the 178 participants recruited for the broader study, 51 (29%) were excluded because they did not return baseline data. Of those 127 patients, 40 (31%) were excluded because they had unknown primary tumor sites, tumors involving the larynx and/or hypopharynx, or tumors involving multiple sites. Eighty-seven (69%) of the 127 patients had a confirmed first diagnosis of either oral or oropharyngeal cancer and were included in this study. Participants with primary oral cancer commonly had involvement of the floor of mouth and/or anterior tongue (54%), but additional subsites (e.g., alveolar ridge, maxilla, mandible) also were represented. Participants with primary oropharyngeal cancer commonly had involvement of base of tongue (54%), along with other subsites (e.g., hypopharynx, posterior pharynx). The heterogeneity of subsites represented within the oral and oropharyngeal cancer groups precluded the use of subsite as a variable in any subsequent statistical analyses.
Procedure
Participants were screened for preliminary eligibility using medical records. Potential participants were approached prior to a pre-operative and/or radiation simulation visit to confirm eligibility criteria were met. Data were collected from questionnaires and the medical record after consent was obtained. PROMs were administered using paper forms and participants completed these forms prior to treatment. Participants who did not return the questionnaires within 3 weeks were contacted once for follow-up. They were paid $20.00 for completion of questionnaires.
Measures
Demographic
Descriptive statistics were used to summarize age, sex, race/ethnicity, marital status, educational achievement, living situation, and employment status. All data were obtained using questionnaires.
Medical
Measures abstracted from the patient medical records included tumor site, tumor stage, and p16 status.
Perceived depression
Perceived depression was measured using the depression subscale of the Hospital Anxiety and Depression Scale (HADS-D).28 The HADS-D includes 7 items rating psychological distress; higher scores indicate higher burden. Summary scores range from 0 (low levels of depression) to 21 (high levels). A cutoff score of 8 is sensitive for depression.
Swallowing-related QOL
The M.D. Anderson Dysphagia Inventory (MDADI)29 was used to evaluate the impact of dysphagia on QOL. It includes 20 items that are averaged to derive a composite score (0 = lowest function; 100 = highest function). A 10 point difference in composite scores is considered clinically meaningful.30
Communicative function
Communicative participation was measured using the CPIB General short form.15,16 The CPIB has been validated in long-term HNC survivors, but it has not been studied in pre-treatment patients. The CPIB General short form includes 10 items that require participants to rate how much their condition (HNC) interferes with communication in everyday settings. Summary scores are converted to T scores (mean = 50, SD = 10), and range from 24.2 – 71.0 (optimal score = 71).
HNC-specific QOL
The University of Washington Quality of Life version 4 (UW-QOL v4)31,32 was used to measure HNC-specific QOL. The UW-QOL includes 12 domain items, 3 global items, and a rating of the most important domains perceived by the patient. Responses for the 12 domains are converted to a numerical score between 0 and 100 (100 = best response), and a composite score may be calculated using the average across domain scores. For this study, only UW-QOL composite scores are reported. A clinically significant change score is 6–7 points.27
Analysis
Missing data analysis was accomplished using SPSS.33 Based on this analysis, missing data appeared to be missing at random. If there was only a single missing value on any measure, missing person-mean imputation34 was used to replace these single missing values. Descriptive statistics were calculated for all categorical variables using frequency distributions and percentages. Means and medians are reported for quantitative measures that were not normally distributed. A comparison of CPIB scores in patients with oral and oropharyngeal cancers with a level of significance set to P < 0.05 was accomplished using a two-sided independent t test.
To determine the independent contribution of communicative participation on UW-QOL scores, controlling for other covariates, we performed a multiple regression with sequential hierarchical predictor entry. If participants were missing data from the primary outcome measure or predictors) participants (N=5) were excluded from the regression analysis. Therefore, 82 participants were included in the final regression analysis model. Variables that were hypothesized to have predictive value were entered into a multiple linear regression model predicting UW-QOL composite scores. Model 1 included demographic covariates: age and sex. Model 2 also included HNC medical covariates: T stage, tumor site (oral versus oropharyngeal), and p16 status. Model 3 also included one psychosocial factor: HADS-D. Finally, model 4 also included swallowing-related QOL and communication measures: MDADI and CPIB General short form scores. Age was retained as a continuous variable (years); sex, tumor site, and p16 status were effect coded. Standard scores were derived for the HADS-D, MDADI, and CPIB. Blocks of variables were reported as changes in R2. Partial correlations and effect sizes (sr2) demonstrated the unique contribution of each variable to UW-QOL composite scores, all other variables being held constant. Standard errors were reported as measures of precision. Pearson correlation coefficients were used to examine relationships between predictors and predicted outcome variables to determine if there was evidence for suppression, mediation, or confounders in the regression model. The significance level for the regression model was set at P < .05. All statistical analyses were conducted using SPSS for Macintosh Version 25.33
Results
Descriptive statistics were provided for 87 participants. More participants were diagnosed with cancer of the oropharynx (N = 52) compared to the oral cavity (N = 35). Most participants were > 50 years of age (N = 74; 85%), male (N = 61; 70%), white (N = 65; 77%) and had received at least some education post high school (N = 52; 62%). Table 1 summarizes additional characteristics. Means and medians at baseline for UW-QOL, CPIB, HADS-D, and composite MDADI scores are provided in Table 2. Median scores consistently demonstrated more dysfunction in patients with oral cancer versus oropharyngeal cancer. There also was a significant difference in CPIB scores between patients with oral (M = 56.68, SD = 13.09) versus oropharyngeal (M = 66.90, SD = 8.85) tumors prior to treatment t(55) = −4.03, (95% CI −15.30 to −5.14; P < .001; adjusted degrees of freedom = 55).
Table I.
Participant Demographic Information
| Characteristic | Oral Cancer | Oropharyngeal Cancer | Total | |||
|---|---|---|---|---|---|---|
| N= 35 | N = 52 | N = 87 | ||||
| N | % | N | % | N | % | |
| Age | ||||||
| 18–34 | 1 | 3% | 0 | 0% | 1 | 1% |
| 35–50 | 5 | 14% | 7 | 14% | 12 | 14% |
| 51–65 | 11 | 31% | 27 | 52% | 38 | 44% |
| 65+ | 18 | 49% | 18 | 35% | 36 | 41% |
| Sex | ||||||
| Male | 20 | 57% | 41 | 79% | 61 | 70% |
| Female | 15 | 43% | 11 | 21% | 26 | 30% |
| Race | ||||||
| White | 27 | 77% | 40 | 77% | 67 | 77% |
| Non-White | 5 | 14% | 8 | 15% | 13 | 15% |
| Unknown | 3 | 9% | 4 | 8% | 7 | 8% |
| Education | ||||||
| Some HS | 1 | 3% | 0 | 0% | 1 | <1% |
| HS Graduate | 14 | 40% | 15 | 29% | 29 | 33% |
| College Graduate | 6 | 17% | 22 | 42% | 28 | 32% |
| Post-College | 11 | 31% | 9 | 17% | 20 | 23% |
| Unknown | 3 | 9% | 6 | 12% | 9 | 10% |
| Living Situation | ||||||
| Live alone | 14 | 40% | 9 | 17% | 23 | 26% |
| Lives with others | 18 | 51% | 36 | 69% | 54 | 62% |
| Unknown | 3 | 9% | 7 | 13% | 10 | 11% |
| Employment Status | ||||||
| Employed FT | 4 | 11% | 20 | 37% | 24 | 34% |
| Employed PT | 4 | 11% | 0 | 0% | 4 | 5% |
| Retired | 15 | 43% | 15 | 29% | 30 | 34% |
| Unemployed | 9 | 26% | 8 | 15% | 17 | 20% |
| Other employment | 0 | 0% | 2 | 4% | 2 | 2% |
| Unknown | 3 | 9% | 6 | 12% | 9 | 10% |
| T stage | ||||||
| T1 or T2 | 17 | 49% | 41 | 79% | 58 | 67% |
| T3 or T4 | 18 | 51% | 11 | 21% | 29 | 33% |
| p16 status | ||||||
| Positive | 0 | 0% | 42 | 81% | 42 | 48% |
| Negative | 33 | 99% | 10 | 19% | 43 | 49% |
| Unknown | 2 | 1% | 0 | 0% | 2 | 2% |
HS= High School; FT= Full Time; PT=Part Time; Unknown=missing data and/or chose not to answer
Table II.
Means and medians at baseline for UW-QOL composite, HADS-D, CPIB General short form, and MDADI composite scores
| Measures | Oral Cancer | Oropharyngeal Cancer | Total | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Mean | (SD) | Median | N | M | (SD) | Median | N | M | (SD) | Median | |
| UW-QOL | 35 | 69.14 | (18.12) | 67.83 | 52 | 86.31 | (12.30) | 90.92 | 87 | 79.40 | (17.07) | 84.00 |
| HADS-D | 35 | 4.66 | (3.87) | 3.00 | 42 | 3.02 | (3.78) | 1.00 | 87 | 3.68 | (3.88) | 2.00 |
| CPIB | 35 | 56.67 | (13.09) | 54.00 | 51 | 66.89 | (8.84) | 71.00 | 86 | 62.74 | (11.84) | 71.00 |
| MDADI | 33 | 70.52 | (8.85) | 68.42 | 50 | 88.04 | (14.59) | 68.42 | 83 | 81.07 | (19.14) | 91.58 |
A correlational analysis of outcome and predictor variables is shown in Table 3. Significant Pearson’s correlations were found between CPIB scores and several other variables: UW-QOL composite scores, T stage, p16 status, tumor site, perceived depression, and swallowing-related QOL, r = 0.37–0.72, P < 0.001. Among predictor variables, significant bivariate correlations also were observed between baseline UW-QOL scores and T stage, p16 status, tumor site, perceived depression, swallowing-related QOL, and communicative participation (see Table 3).
Table III.
Correlation Table for Linear Regression using Sequential Predictor Entry
| Measure | M | (SD) | 1. | 2. | 3. | 4. | 5. | 6. | 7. | 8. | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Outcomes | |||||||||||||||||
| 1. UW-QOL Compositi | 79.25 | (16.78) -- | |||||||||||||||
| Block 1 Demographic | |||||||||||||||||
| 2. Age | 62.13 | (11.55) | −.02 | -- | |||||||||||||
| 3. Sex | 0.42 | (0.92) | .17 | −.10 | -- | ||||||||||||
| Block 2 Medical | |||||||||||||||||
| 4. T stage | 2.27 | −(1.01) | −.40 *** | .04 | −.27 ** | -- | |||||||||||
| 5. p16 status | 0.00 | (1.00) | .37 *** | −.09 | .32 ** | −.32 ** | -- | ||||||||||
| 6. Tumor site | −0.20 | (0.99) | −.49 *** | .08 | −.24 * | .35 ** | −.82 *** | -- | |||||||||
| Block 3 Psychosocial | |||||||||||||||||
| 7. HADS Depression Score | |||||||||||||||||
| 3.71 | (3.94) | −.64 *** | −.64 | −.21 * | .17 | −.12 | .21 * | -- | |||||||||
| Block 4 Functional | |||||||||||||||||
| Speech and Swallowing | |||||||||||||||||
| Measures | |||||||||||||||||
| 8. MDADI Composite | 80.95 | (19.22) | .71 *** | .00 | .12 | −.45 *** | .38 *** | −.45 *** | −.46 *** | -- | |||||||
| 9. CPIB T Score | 62.78 | (11.75) | .72 *** | .15 | .09 | −.40 *** | .39 *** | −.42 *** | −.48 *** | .75 *** | |||||||
Note. N = 82. UWQOL = University of Washington Quality of Life; HADS = Hospital Anxiety and Depression Scale; MDADI = MD Anderson Depression Inventory; CPIB-10 = Communication Participation Item Bank-10
asterisks reflect the range of p values for each correlation (* p < 0.05, ** p < 0.01, *** p < 0.001).
Analysis of covariates as predictors of UW-QOL included data from 82 participants. Results from the final block with all predictors entered into the model showed that the final model explained 71% of the variance in UW-QOL scores (see Table 4). Tumor site, T stage, and p16 status accounted for 28% of variance in UW-QOL. Depression scores predicted an additional 28% of the variance in QOL scores. Swallowing-related QOL and communicative participation together predicted an additional 12% of variance in UW-QOL. When all predictors were entered into the final model, tumor site, perceived depression, swallowing-related QOL, and communicative participation were unique predictors of UW-QOL, but demographic variables, T stage, and p16 status were not, holding other covariates constant. Participants with oral cancers had UW-QOL scores that were 8.4 points lower than those with oropharyngeal cancer, holding all else constant. For every 1 SD increase in CPIB scores at baseline, participants had UW-QOL scores that were 5.1 points higher, holding all else constant (see Table 4).
Table IV.
Multiple Linear Regression with Sequential Predictor Entry for UW-QOL
| Block 1 | Block 2 | Block 3 | Block 4 | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| R2change | R2total | R2adj | b | sr2 | R2change | R2total | R2adj | b | sr2 | R2change | R2total | R2adj | b | sr2 | R2change | R2total | R2adj | b | sr2 | |
| Model Fit | 0.03 | 0.03 | 0.00 | 0.28 *** | 0.31 *** | 0.26 | 0.28 *** | 0.59 *** | 0.56 | 0.12 *** | 0.71 *** | 0.68 | ||||||||
| Coefficients | ||||||||||||||||||||
| Intercept | 78.03 *** | 83.77 *** | 88.46 | 87.12 *** | ||||||||||||||||
| Age | 0.00 | 0.00 | 0.04 | 0.00 | −0.04 | 0.00 | −0.10 | 0.00 | ||||||||||||
| Sex | 3.09 | 0.03 | 0.43 | 0.00 | −1.58 | 0.01 | −0.19 | 0.00 | ||||||||||||
| T stage | −3.76 * | 0.07 | −3.11 * | 0.05 | −0.94 | 0.00 | ||||||||||||||
| p16 status | −2.17 | 0.02 | 0.03 | 0.00 | −1.73 | 0.01 | ||||||||||||||
| Primary site of lesion | −8.54 ** | 0.25 | −5.36 * | 0.10 | −4.22 * | 0.06 | ||||||||||||||
| HADS Depression | −9.25 *** | 0.31 | 0.12 | |||||||||||||||||
| MDADI Score | 3.77 * | 0.05 | ||||||||||||||||||
| CPIB T Scores | 5.07 ** | 0.03 | ||||||||||||||||||
Note. N =82. Block 1 F-change test df = 2,79; Block 2 df = 3, 76; Block 3 df = 1, 75, Block 4 df =2, 73. UWQOL = University of Washington Quality of Life; HADS = Hospital Anxiety and Depression Scale; MDADI = MD Anderson Depression Inventory; CPIB = Communication Participation Item Bank
asterisks reflect the range of p values for each predictor (* p < 0.05, ** p < 0.01, *** p < 0.001).
Discussion
Results from this study add to the extant literature documenting pre-treatment measures of communicative function in patients with oral and oropharyngeal HNC. Median and mean baseline CPIB General short form scores (median T score = 71.0; mean T score = 62.74) for our overall sample were 1 to 2 standard deviations higher than average CPIB T scores previously reported in long-term HNC survivors (Mean T scores = 49.9 – 53.0).19,35 Thus, as a group, few limitations were reported in everyday communication prior to treatment. However, findings are more revealing when communicative participation is examined with regard to other variables at baseline. Importantly, CPIB scores in patients with oral cancer were significantly worse than those with oropharyngeal cancer (mean T scores = 56.67 vs. 66.89, respectively; P < 0.001). About half of all oral cancer patients reported communication difficulty prior to treatment, consistent with one of the few longitudinal studies documenting speech-related QOL outcomes in oral cancer patients.7
Tumor site was strongly related to p16 status because all subjects included in this analysis with primary tumors of the oral cavity were p16 negative. Therefore, it is difficult to extricate the effect of site of lesion from Human Papilloma Virus (HPV)-related status on communication. Regardless, based on previously reported data in other long-term HNC survivors,19,35 it appears that patients with oropharyngeal cancers may be at high risk for not returning to baseline in communication function, even many years after treatment. It also appears that a large number of patients with oral cancer report difficulties both pre- and post-treatment. Pre-treatment CPIB scores from participants in this study support the long-held contention that HNC treatments such as surgery and (chemo)radiation therapy may have long-lasting, negative effects on communication.1 Future studies using longitudinal designs will help strengthen the interpretation of these findings.
The median UW-QOL composite score for participants in this study was 84 (SD = 17.07; range: 24.71 – 100), similar to previously reported pre-treatment scores in oral and oropharyngeal HNC patients.26 Participants with high risk HPV-related cancers did not have significantly higher pre-treatment QOL when controlling for tumor site in the multivariate analysis. However, higher baseline UW-QOL scores were found in patients with oropharyngeal cancer than those with oral cancer (medians = 90.92 vs. 67.83, respectively), which is a clinically significant difference, consistent with previous findings.27 Higher baseline UW-QOL scores also were associated with lower tumor stages.26
An important result of this study relates to the predictive strength of the variables entered into the final regression model. Specifically, 71% of the variance in UW-QOL scores was predicted, with tumor site, depression, swallowing-related QOL, and communication measures uniquely predicting UW-QOL scores, holding all else constant in the final model. Unsurprisingly, medical factors such as tumor site, and psychosocial factors, such as perceived depression, were strong predictors of QOL.3,25,26 Results also showed that 12% of the variance in baseline UW-QOL was predicted by swallowing-related QOL and communication measures, above and beyond other variables. These findings are consistent with others who have documented the importance of swallowing in this population.36 Communication and swallowing related QOL were highly correlated, which was expected in this patient population. Despite the high level of correlation between these measures, the CPIB scores uniquely predicted a significant amount of variance above and beyond that of all other predictor variables. The relationship between communication and UW-QOL scores is consistent with findings reported previously in long-term HNC survivors.16 The results highlight the importance of considering communication outcomes in oral and oropharyngeal HNC patients, even at baseline.3
The ability to measure the impact of HNC and its treatment on communication in everyday settings is of paramount importance as younger individuals are increasingly diagnosed with high-risk HPV-related HNC cancers. This population is highly vulnerable to restrictions in communicative participation because they are engaged in a variety of activities and will live longer with the consequences of HNC and treatment toxicities.26 One substantial benefit of using a PROM such as the CPIB above others, such as the Speech Handicap Index,10 is that the CPIB has been validated using Item Response Theory. Using an IRT-based PROM can enhance precision and reliability, facilitates comparisons across centers and research studies, and eases its use in meta-analyses and systematic reviews that provide a basis for standard of care. Recent development of computerized adaptive testing for the CPIB further reduces response burden, important for patients and clinical providers alike.37 It also facilitates the use of such a measure in clinical and research practice, which is often over-looked in this patient population.
Results from this study should be considered in light of the participant sample and other limitations with study design. First, patients who completed questionnaires were well enough to respond. Most participants were white and reported a high level of education. Consequently, our results may not be representative of more diverse HNC populations. Additionally, the tumor subsite location might also inform the impact of communicative participation on QOL. However, only tumor site (oral versus oropharyngeal) was controlled for in our analysis to optimize statistical power. As with other exploratory studies, causal relations between predictor variables and QOL cannot be made. Finally, this study was limited to measurement at a single time point (pre-treatment). Prospective follow up will be important for interpreting results and suggesting directions for rehabilitation.
Conclusions
The CPIB is a sensitive PROM for measuring functional speech-related difficulties in patients with oral and oropharyngeal HNC, which is important for QOL even at pre-treatment.21 Future studies should include valid pre-treatment measures that consider communication for differentiating the effect of the tumor from treatment effects, comparing the efficacy of different treatment modalities, and tracking changes in function over time.
Acknowledgements:
We gratefully acknowledge funding support from National Institutes of Health / National Cancer Institute (R01CA177635; principal investigator, Tanya Eadie), the Vocal Function Lab of the University of Washington Department of Speech and Hearing Sciences, study participants, and the clinicians at the University of Washington Medical Center that assisted with recruitment. We also acknowledge the generous support and contributions of Dr. Eduardo Méndez.
Footnotes
Portions of this paper were presented at the World Congress of the International Academy of Oral Oncology, August 31 – September 3, 2019, Rome, Italy.
Conflicts of interest: None
References
- 1.Dwivedi RC, St. Rose S, Chisholm EJ, et al. Evaluation of speech outcomes using English version of the Speech Handicap Index in a cohort of head and neck cancer patients. Oral Oncology. 2012;48(6):547–553. [DOI] [PubMed] [Google Scholar]
- 2.Dwivedi RC, Kazi RA, Agrawal N, et al. Evaluation of speech outcomes following treatment of oral and oropharyngeal cancers. Cancer Treatment Reviews. 2009;35(5):417–424. [DOI] [PubMed] [Google Scholar]
- 3.Karnell LH, Funk GF, Hoffman HT. Assessing head and neck cancer patient outcome domains. Head Neck. 2000;22(1):6–11. [DOI] [PubMed] [Google Scholar]
- 4.Jacobi I, van der Molen L, Huiskens H, van Rossum MA, Hilgers FJM. Voice and speech outcomes of chemoradiation for advanced head and neck cancer: a systematic review. Eur Arch Otorhinolaryngol. 2010;267(10):1495–1505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.van der Molen L, van Rossum MA, Burkhead LM, Smeele LE, Rasch CRN, Hilgers FJM. A Randomized Preventive Rehabilitation Trial in Advanced Head and Neck Cancer Patients Treated with Chemoradiotherapy: Feasibility, Compliance, and Short-term Effects. Dysphagia. 2011;26(2):155–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rinkel RN, Verdonck-de Leeuw IM, Doornaert P, et al. Prevalence of swallowing and speech problems in daily life after chemoradiation for head and neck cancer based on cut-off scores of the patient-reported outcome measures SWAL-QOL and SHI. European Archives of Oto-Rhino-Laryngology. 2016;273(7):1849–1855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Head and Neck Research Network, Dzioba A, Aalto D, et al. Functional and quality of life outcomes after partial glossectomy: a multi-institutional longitudinal study of the head and neck research network. J of Otolaryngol - Head & Neck Surg. 2017;46(1):56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Eadie TL, Yorkston KM, Klasner ER, et al. Measuring Communicative Participation: A Review of Self-Report Instruments in Speech-Language Pathology. Am J Speech Lang Pathol. 2006;15(4):307–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Op de Coul BMR, Ackerstaff AH, Van As CJ, et al. Quality of life assessment in laryngectomized individuals: do we need additions to standard questionnaires in specific clinical research projects?: QoL assessment in laryngectomized individuals. Clinical Otolaryngology. 2005;30(2):169–175. [DOI] [PubMed] [Google Scholar]
- 10.Rinkel RN, Leeuw IMV, van Reij EJ, Aaronson NK, Leemans CR. Speech Handicap Index in patients with oral and pharyngeal cancer: Better understanding of patients’ complaints. Head Neck. 2008;30(7):868–874. [DOI] [PubMed] [Google Scholar]
- 11.Funk GF. Long-term Health-Related Quality of Life in Survivors of Head and Neck Cancer. Arch Otolaryngol Head Neck Surg. 2012;138(2):123. [DOI] [PubMed] [Google Scholar]
- 12.Baylor CR, Yorkston KM, Eadie TL, Miller RM, Amtmann D. Developing the Communicative Participation Item Bank: Rasch Analysis Results From a Spasmodic Dysphonia Sample. Journal of Speech, Language, and Hearing Research. 2009;52(5):1302–1320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Baylor C, Burns M, Eadie T, Britton D, Yorkston K. A Qualitative Study of Interference With Communicative Participation Across Communication Disorders in Adults. Am J Speech Lang Pathol. 2011;20(4):269–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yorkston KM, Baylor CR, Dietz J, et al. Developing a scale of communicative participation: A cognitive interviewing study. Disability and Rehabilitation: Rehabilitation Outcome Measures on the Horizon. 2008;30(6):425–433. [DOI] [PubMed] [Google Scholar]
- 15.Baylor C, Yorkston K, Eadie T, Kim J, Chung H, Amtmann D. The Communicative Participation Item Bank (CPIB): Item Bank Calibration and Development of a Disorder-Generic Short Form. J Speech Lang Hear Res. 2013;56(4):1190–1208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Eadie TL, Lamvik K, Baylor CR, Yorkston KM, Kim J, Amtmann D. Communicative Participation and Quality of Life in Head and Neck Cancer. Ann Otol Rhinol Laryngol. 2014;123(4):257–264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Eadie TL, Otero D, Cox S, et al. The relationship between communicative participation and postlaryngectomy speech outcomes: Postlaryngectomy speech outcomes. Head Neck. 2016;38(S1):E1955–E1961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Eadie T, Faust L, Bolt S, et al. Role of Psychosocial Factors on Communicative Participation among Survivors of Head and Neck Cancer. Otolaryngol Head Neck Surg. 2018;159(2):266–273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bolt S, Eadie T, Yorkston K, Baylor C, Amtmann D. Variables Associated With Communicative Participation After Head and Neck Cancer. JAMA Otolaryngol Head Neck Surg. 2016;142(12):1145–1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ringash J, Bernstein LJ, Cella D, et al. Outcomes toolbox for head and neck cancer research: Outcomes Toolbox for Head and Neck Cancer. Wax MK, ed. Head Neck. 2015;37(3):425–439. [DOI] [PubMed] [Google Scholar]
- 21.Borggreven PA, Aaronson NK, Verdonck-de Leeuw IM, et al. Quality of life after surgical treatment for oral and oropharyngeal cancer: a prospective longitudinal assessment of patients reconstructed by a microvascular flap. Oral Oncol. 2007;43(10):1034–1042. [DOI] [PubMed] [Google Scholar]
- 22.Lango MN, Egleston B, Fang C, et al. Baseline health perceptions, dysphagia, and survival in patients with head and neck cancer. Cancer. 2014;120(6):840–847. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kalavrezos N, Cotrufo S, Govender R, et al. Factors affecting swallow outcome following treatment for advanced oral and oropharyngeal malignancies: Factors affecting swallow outcome. Head Neck. 2014;36(1):47–54. [DOI] [PubMed] [Google Scholar]
- 24.Chan JYK, Lua LL, Starmer HH, Sun DQ, Rosenblatt ES, Gourin CG. The relationship between depressive symptoms and initial quality of life and function in head and neck cancer. The Laryngoscope. 2011;121(6):1212–1218. [DOI] [PubMed] [Google Scholar]
- 25.Howren MB, Christensen AJ, Hynds Karnell L, Van Liew JR, Funk GF. Influence of pretreatment social support on health-related quality of life in head and neck cancer survivors: Results from a prospective study. Head Neck. 2013;35(6):779–787. [DOI] [PubMed] [Google Scholar]
- 26.Sharma A, Méndez E, Yueh B, et al. Human Papillomavirus–Positive Oral Cavity and Oropharyngeal Cancer Patients Do Not Have Better Quality-of-Life Trajectories. Otolaryngol Head Neck Surg. 2012;146(5):739–745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.El-Deiry MW, Futran ND, McDowell JA, Weymuller EA, Yueh B. Influences and Predictors of Long-term Quality of Life in Head and Neck Cancer Survivors. Arch Otolaryngol Head Neck Surg. 2009;135(4):380. [DOI] [PubMed] [Google Scholar]
- 28.Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67(6):361–370. [DOI] [PubMed] [Google Scholar]
- 29.Chen AY, Frankowski R, Bishop-Leone J, et al. The development and validation of a dysphagia-specific quality-of-life questionnaire for patients with head and neck cancer: the M. D. Anderson dysphagia inventory. Arch Otolaryngol Head Neck Surg. 2001;127(7):870–876. [PubMed] [Google Scholar]
- 30.Hutcheson KA, Barrow MP, Lisec A, Barringer DA, Gries K, Lewin JS. What is a clinically relevant difference in MDADI scores between groups of head and neck cancer patients?: Clinically Meaningful Difference in MDADI Scores. The Laryngoscope. 2016;126(5):1108–1113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Weymuller EAJ, Yueh B, Deleyiannis FW, Kuntz AL, Alsarraf R, Coltrera MD. Quality of life in head and neck cancer. Laryngoscope. 2000;110(3 Pt 3):4–7. [DOI] [PubMed] [Google Scholar]
- 32.Rogers SN, Lowe D, Brown JS, Vaughan ED. The University of Washington head and neck cancer measure as a predictor of outcome following primary surgery for oral cancer. Head Neck. 1999;21(5):394–401. [DOI] [PubMed] [Google Scholar]
- 33.IBM. SPSS Statistics for Windows, Version 25.0 Armonk, NY: IBM Corp.; 2017. [Google Scholar]
- 34.Bernaards CA, Sijtsma K. Influence of Imputation and EM Methods on Factor Analysis when Item Nonresponse in Questionnaire Data is Nonignorable. Multivariate Behav Res. 2000;35(3):321–364. [DOI] [PubMed] [Google Scholar]
- 35.Eadie T, Kapsner-Smith M, Bolt S, Sauder C, Yorkston K, Baylor C. Relationship between perceived social support and patient-reported communication outcomes across communication disorders: a systematic review. Int J Lang Commun Disord. Published online July 24, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Nguyen NP, Frank C, Moltz CC, et al. Impact of dysphagia on quality of life after treatment of head-and-neck cancer. Int J Radiat Oncol Biol Phys. 2005;61(3):772–778. [DOI] [PubMed] [Google Scholar]
- 37.Ringash J, Thariat J. Improving Head and Neck Cancer Outcomes: Technology, Used Wisely. Int J Radiat Oncol Biol Phys. 2016;96(3):489–492. [DOI] [PubMed] [Google Scholar]