Abstract
Background:
The Questionnaire of Olfactory Disorders - Negative Statements (QOD-NS) is an important instrument in the measurement of olfactory-specific quality of life (QOL). In the clinical setting, patients can be overwhelmed with the time required to complete questionnaires. Our objective was to develop a brief version of the QOD-NS to streamline clinical care and research.
Methods:
QOD-NS scores from 221 subjects were used to determine what subset of the seventeen QOD-NS questions best correlate with total and subdomain QOD-NS scores. An initial pool of 11 questions was made by removing items with ρ < 0.80 to their respective subdomain scores. Next, 500 bootstrapped samples were taken and on each, an all subsets regression was performed with total QOD-NS scores and QOD-NS subdomain scores as the outcomes. From that, our “top” and “bottom” ten subsets were identified based on mean r2, representation in bootstrap analysis, and number of items.
Results:
All of our top subsets had excellent correlation with total and subdomain QOD-NS scores (mean r2 > 0.90). Our top choice has 7 total questions, is representative of all subdomains, has mean r2 of 0.92, and was represented in 323 of our 500 bootstrapped samples. The worst-performing subset has 5 items, mean r2 0.81, and was represented in only one bootstrapped sample.
Conclusions:
Using less than half of the questions in the QOD-NS, excellent correlations with both total and domain-specific scores are achieved. A brief version of the QOD-NS may prove useful in future clinical and research settings.
Keywords: olfaction, quality of life, questionnaire of olfactory disorders, smell, sinusitis
INTRODUCTION
Chronic rhinosinusitis (CRS) is a common inflammatory disorder affecting approximately 5–16% of North American populations.1 Symptoms of CRS are nasal congestion, discolored nasal drainage, facial pain/pressure and olfactory dysfunction (OD), with OD estimated to affect between 40 to 80% of CRS patients in certain study populations.2 Known impacts of OD include economic influences, decreased productivity, and declined quality of life (QOL).3
Recent research has assessed patient outcomes following endoscopic sinus surgery (ESS) for CRS. Commonly used instruments include disease-specific QOL questionnaires like the 22-item Sinonasal Outcome Test (SNOT-22)4. However, these instruments do not focus on smell or taste, and usually have only a single question devoted for olfaction. This has led to the development of validated questionnaires that focus on olfactory-specific quality of life (QOL).5–7 Specifically, the questionnaire of olfactory disorders negative statements (QOD-NS) has proven to be an important instrument in the measurement of olfactory-specific QOL.8 The QOD-NS has been shown to have good psychometric validity9 and has been shown to correlate with objective olfactory loss.10 Our prior work has also identified several subdomains within the QOD-NS. Subdomains are subsets of questions within the instrument that have common overarching themes. In the QOD-NS, there are subdomains of questions that focus around social, eating, anxiety, and annoyance factors.8 However, the length of this 17-item questionnaire can be problematic. In the clinical and research settings, patients can be overwhelmed with the time required to complete questionnaires for both clinical practice evaluations and investigational protocols, and it has been shown that shortening the length of questionnaires can lead to an increase in response rates.11 Furthermore, shortening patient reported outcome metrics (PROMs) can lead to increased efficiency, reduced questionnaire burden, and increase the quality of collected data.12 As such, shortening the QOD-NS is an appealing strategy to decrease the burden imposed on patients, increase the likelihood of questionnaire completion, and maximize the accuracy and quality of responses in future clinical research on olfactory-specific QOL.
The current study has the following objectives: 1) develop a brief version of the QOD-NS to streamline clinical care and research without impacting the accuracy of measuring olfactory-specific QOL measurement; and 2) to ensure the shortened version of the QOD-NS is representative of all subdomains of the original questionnaire.
METHODS
Study Population
Adult participants (≥18 years of age) with a diagnosis of CRS as defined by the American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) were prospectively enrolled.13 Participants were enrolled between August, 2012, and June, 2015, into a nonrandomized, multicenter North American cohort studying the impacts of ESS and medical therapy on CRS outcomes. IRB approval as obtained at each participating site, and the participants provided researchers written consent at enrollment meetings in various sinus clinics in North American, academic hospitals (Oregon Health & Science University, Stanford University, the Medical University of South Carolina, and the University of Calgary).
Questionnaire of Olfactory Disorders-Negative Statements
The primary objectives in this study were to shorten the QOD-NS while maintaining the accuracy of the new shortened instrument. The original QOD-NS consists of 17 statements which patients reported on a scale of 0 to 3, with higher scores reflecting better olfactory-specific QOL.14 The original QOD-NS results in a score from 0 to 51 given the sum of graded responses from each statement. Furthermore, we have reported on sub-domains which exist within the QOD-NS, and these were used in the current study.8
Statistical Analysis
Baseline QOD-NS scores were used to determine what subset of the seventeen QOD-NS questions best correlate with total and subdomain QOD-NS scores. First, in order to reduce computational exertion, an initial subset of items was made. Items with ρ < 0.80 with their respective subdomain were removed from the analysis. Next, a bootstrapping analysis was performed. This technique, in effect, leverages qualities of the bootstrapping technique whereby many simple random samples are taken and independently evaluated in order to more accurately approximate the general population. In addition, the variability between the bootstrapped samples provides the ability to estimate the variability of each of the 5 correlations. 500 bootstrapped samples were taken and on each, an all subsets regression was performed with total QOD-NS scores and QOD-NS subdomain scores as the outcomes.15 Regression analysis was used to ensure questions chosen in the shortened QOD-NS positively correlate with the data of the original QOD-NS. From the all subsets regression, the 50 subsets with the best r2 values which also had representation from all 4 subdomains were retained. For each remaining subset, regression analyses were performed for each subdomain. For each bootstrapped sample and each of the 50 retained subsets, five correlations (total QOD and one for each of the four QOD subdomains) and their mean were computed and stored for analyses.
Summary statistics for each subset were computed to determine the number of times each subset was selected from the individual bootstrap analyses and the distribution of the correlations. To select the ‘top ten’ subsets, a requirement of nine or fewer items and an N of ≥ 250 had to be met. From that, subsets were identified starting with the smallest number of items and highest average mean correlation (mean of the five correlations) until 10 subsets had been chosen. To select the ‘bottom ten’ subsets, a requirement of N ≤ 10 had to be met and the lowest ten average mean correlations were chosen. All analyses were performed using SAS v9.4 ©.
RESULTS
In this cohort, 221 subjects completed QOD-NS questionnaires at baseline. Baseline patient demographic and clinical data are presented in Table 1 as mean ± standard deviation and frequency (%) for continuous and categorical variables, respectively. Where appropriate, continuous variables are presented as median (IQR).
Table 1:
Patient Characteristics at Baseline
| Demographic/QOL Measure | Mean (SD) Frequency (%) Median (IQR) | |
|---|---|---|
| Age | 49.8 (15.9) | |
| Sex | Male | 101 (45.7) |
| Female | 120 (54.3) | |
| Race | White | 192 (86.9) |
| Non-White | 29 (13.1) | |
| Ethnicity | Hispanic | 11 (5) |
| Non-Hispanic | 210 (95) | |
| Asthma | 97 (43.9) | |
| Smoking history | 10 (4.6) | |
| CRS Measures | ||
| Polyp status | CRSwNP | 83 (37.6) |
| CRSsNP | 138 (62.4) | |
| SNOT-22 Score | 53.3 (21.5) | |
| Endoscopy score | 5.5 (3.8) | |
| CT score | 11.6 (6.3) | |
| Olfaction QOL Measures | ||
| QOD-NS Total | 17.0 (IQR 9.0, 28.0) | |
| QOD-NS Subdomain 1 | 0.17 (IQR 0.0, 0.8) | |
| QOD-NS Subdomain 2 | 0.75 (IQR 0.0, 1.5) | |
| QOD-NS Subdomain 3 | 0.25 (IQR 0.0, 1.0) | |
| QOD-NS Subdomain 4 | 1.33 (IQR 0.7, 2.3) | |
CRS – Chronic Rhinosinusitis, AFS – Allergic Fungal Sinusitis, SF-6D – Short Form 6D, CT – computed tomography, SNOT-22 – 22 item Sinonasal Outcome Test, QOD-NS – Questionnaire of Olfactory Disorders Negative Statements
Our preliminary subset of questions consisted of 11 items that had correlations > 0.80 with their respective subdomains. This allowed for reduction of computational exertion from calculating more than half a million (219 = 524,288) correlations 500 times. Our preliminary subset of 11 questions is detailed in Table 2, and was then used through the rest of our analysis. Bootstrap analysis results are detailed in Table 3 and shown in Figure 1. All of our top subsets had excellent correlation with total and subdomain QOD-NS scores (mean r2 > 0.90). Our top choice has 7 total questions, is representative of all subdomains, has mean r2 of 0.92, and was represented in 323 of our 500 bootstrapped samples. The worst-performing subset has 5 items, mean r2 0.81, and was represented in only one bootstrapped sample. The statistical outputs of the subset items were then analyzed and are shown in Table 4, also represented in Figure 1. In Figure 1, the top performing subsets are represented in blue with lowest performing subsets in red. The detailed questions included in our top-performing subset, and the relative representation of each QOD-NS subdomain are shown in Table 5.
Table 2:
Preliminary Question Pool for Bootstrap Analysis
| Factor 1: Social factor | |
| Question # | Question text |
| 33 | The changes in my sense of smell make me feel isolated. |
| 42 | Because of the changes in my sense of smell I have problems with taking part in activities of daily life. |
| 49 | The changes in my sense of smell make me feel angry. |
| Factor 2: Eating factor | |
| Question # | Question text |
| 1 | Because of the changes in my sense of smell, I go to restaurants less often than I used to. |
| 11 | Because of the changes in my sense of smell, I don’t enjoy drinks or food as much as I used to. |
| 37 | Because of the changes in my sense of smell I eat less than I used to or more than I used to. |
| Factor 3: Anxiety factor | |
| Question # | Question text |
| 15 | Because of the changes in my sense of smell, I feel more anxious than I used to feel. |
| 26 | Because of the changes in my sense of smell I visit friends, relatives, or neighbors less often. |
| 27 | Because of the changes in my sense of smell, I try harder to relax. |
| Factor 4: Annoyance factor | |
| Question # | Question text |
| 13 | I am worried that I will never get used to the changes in my sense of smell. |
| 22 | The changes in my sense of smell annoy me when I am eating. |
The 11 items from the original 17 QOD-NS items that were identified to have high correlation with respective subdomains for bootstrap analysis. These items had a p > 0.80 while the 5 removed items had p < 0.80 (not shown here).
Table 3:
Top and Bottom Performing Subsets of QOD-NS Questions
| Subset | QOD-NS Questions | Number of Items Per Subset | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Social Questions (Factor 1) |
Eating Questions (Factors 2) |
Anxiety Questions (Factor 3) |
Annoyance Questions (Factor 4) |
|||||||||
| 33 | 42 | 49 | 1 | 11 | 37 | 15 | 26 | 27 | 22 | 13 | ||
| Top Performing Subsets | ||||||||||||
| A | 1 | 1 | 1 | 1 | 0 | 1 | 0 | 0 | 1 | 0 | 1 | 7 |
| B | 0 | 1 | 1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 7 |
| C | 1 | 1 | 1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 1 | 8 |
| D | 1 | 1 | 1 | 1 | 0 | 1 | 0 | 1 | 1 | 1 | 1 | 9 |
| E | 1 | 1 | 1 | 1 | 0 | 1 | 1 | 0 | 1 | 1 | 1 | 9 |
| F | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | 0 | 1 | 8 |
| G | 1 | 1 | 1 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 1 | 8 |
| H | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | 1 | 1 | 9 |
| I | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 1 | 0 | 1 | 1 | 9 |
| J | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 1 | 0 | 1 | 9 |
| Bottom Performing Subsets | ||||||||||||
| K | 0 | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 1 | 1 | 0 | 4 |
| L | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 1 | 1 | 0 | 5 |
| M | 1 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 5 |
| N | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 4 |
| O | 0 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 5 |
| P | 0 | 0 | 1 | 1 | 0 | 1 | 0 | 0 | 1 | 1 | 0 | 5 |
| Q | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 4 |
| R | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 5 |
| S | 1 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 5 |
| T | 1 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 5 |
Characteristics of the 4 Subdomains from the 11 QOD-NS items. The 11 items were classified into respective subdomains of social, eating, anxiety, and annoyance questions.
Figure 1: Bubble Plot of QOD-NS Question Subsets.
Each bubble represents a subset of QOD-NS questions. Bubble size represents the number of QOD-NS questions in the subset. Blue bubbles represent the “top ten” performing subsets, red bubbles represent the “bottom ten” performing subsets. This figure illustrates the top and bottom subsets determined through correlation with the original questionnaire and bootstrap samples.
Table 4:
Performance Statistics for QOD-NS Subsets
| Subset | Number of Bootstrapped Samples | Mean R-squared | Standard Deviation | Number of Questions in Sample |
|---|---|---|---|---|
| Top Performing Subsets | ||||
| A | 323 | 0.9163 | 0.0079 | 7 |
| B | 281 | 0.9068 | 0.0090 | 7 |
| C | 467 | 0.9157 | 0.0081 | 8 |
| D | 352 | 0.9280 | 0.0064 | 9 |
| E | 415 | 0.9347 | 0.0056 | 9 |
| F | 425 | 0.9210 | 0.0076 | 8 |
| G | 318 | 0.9093 | 0.0086 | 8 |
| H | 488 | 0.9208 | 0.0076 | 9 |
| I | 404 | 0.9241 | 0.0068 | 9 |
| J | 408 | 0.9401 | 0.0050 | 9 |
| Bottom Performing Subsets | ||||
| K | 1 | 0.7895 | n/a | 4 |
| L | 2 | 0.8262 | 0.0089 | 5 |
| M | 2 | 0.8205 | 0.0183 | 5 |
| N | 1 | 0.8212 | n/a | 4 |
| O | 3 | 0.8186 | 0.0150 | 5 |
| P | 3 | 0.8245 | 0.0090 | 5 |
| Q | 6 | 0.7928 | 0.0161 | 4 |
| R | 5 | 0.8208 | 0.0178 | 5 |
| S | 6 | 0.8125 | 0.0077 | 5 |
| T | 1 | 0.8104 | n/a | 5 |
The bootstrap analysis of the top performing and bottom performing subsets are represented here. This table gives the mean r2, SD, and number of questions in each sample that were used in our analysis.
Table 5:
‘Subset A’ Questions, Displayed by QOD-NS Factors
| Factor 1: Social questions | |
| Question # | Question text |
| 33 | The changes in my sense of smell make me feel isolated. |
| 42 | Because of the changes in my sense of smell I have problems with taking part in activities of daily life. |
| 49 | The changes in my sense of smell make me feel angry. |
| Factor 2: Eating questions | |
| Question # | Question text |
| 1 | Because of the changes in my sense of smell, I go to restaurants less often than I used to. |
| 37 | Because of the changes in my sense of smell I eat less than I used to or more than I used to. |
| Factor 3: Anxiety questions | |
| Question # | Question text |
| 27 | Because of the changes in my sense of smell, I try harder to relax. |
| Factor 4: Annoyance questions | |
| Question # | Question text |
| 13 | I am worried that I will never get used to the changes in my sense of smell. |
The items identified as part of the “top subset” are presented here. These items best correlated with the olfactory-specific QOL results from the original QOD-NS.
DISCUSSION
The QOD-NS has been shown to be a robust olfactory-specific QOL instrument that has good psychometric properties, correlates with objective olfactory parameters, and can be used to measure olfactory-specific outcomes in the treatment of CRS.2,9,10 As such, it has become one of the primary instruments utilized in clinical investigations surrounding olfaction in CRS. However, the evolution in the complexity and sophistication of recent CRS clinical research has been associated with a concomitant increase the number of tests and questionnaires that patients are asked to complete. While this results in robust datasets that can provide great insight into the patient condition and the outcomes of our therapies, it can also place a significant burden on patients. Reducing the time required to complete different components of a clinical research protocol by just a few minutes can have a significant cumulative effect, particularly in a busy clinical setting where most of our patients are recruited. Importantly, prior research suggests that brevity in questionnaires can lead to greater response rates and better data quality.11,12
The current study was designed to effectively shorten the length of the QOD-NS while also maintaining consistency in measured patient-reported outcomes of olfactory-specific QOL. We also aimed to keep items from the four different factorial subdomains of the QOD-NS (social, eating, anxiety, and annoyance) that have been previously reported.8 Items of the QOD-NS were assessed through regression to determine which items best correlated with the overall QOL measures from the original QOD-NS and with the subdomain specific to each of the original items of the QOD-NS. Our goal was to find the subset with lowest number of questions that would also produce scores that had excellent correlation with total and domain-specific scores. The 7 items of the top-performing subset did span across all four subdomains of social, eating, anxiety, and annoyance questions. These items included three from the social subdomain involving impacts on patients’ daily activities and feelings of isolation and anger; two from the eating subdomain regarding impacts on attending restaurants and food consumption; one from the anxiety subdomain inquiring on relaxation; and one from the annoyance subdomain involving adapting to changes in olfaction dysfunction. Interestingly, in all of our top-performing subsets, factor 3 (anxiety) was least represented. This may suggest that certain subdomains of the QOD-NS may differentially impact total QOD-NS scores.
This study has several limitations which should be noted. One major limitation is that the participants had high associated disease severity and were assessed through tertiary practices, and therefore these results may not be externally generalizable to all CRS patients. Furthermore, our initial criteria for selection of a best-performing subset (i.e. shortest possible survey with the highest correlation to the original scores) may not necessarily equate to the best possible subset of QOD-NS questions with the highest fidelity to the original instrument. Further validation and study of our shortened version of the QOD-NS will be necessary to confirm its clinical utility and validity.
CONCLUSION
Using less than half of the questions in the QOD-NS, excellent correlations with both total and domain-specific scores are achieved. A brief version of the QOD-NS may prove useful in future clinical and research settings by improving efficiency and data quality, and reducing patient burden.
Supplementary Material
Funding:
This work was supported by grants from the National Institute on Deafness and Other Communication Disorders (NIDCD), one of the National Institutes of Health, Bethesda, MD (R03 DC013651–01; PI: ZM Soler and R01 DC005805; PIs: TL Smith and ZM Soler).
Potential Conflicts of Interest: JLM is a consultant for Sanofi Genzyme, which is not affiliated with this manuscript. RJS is consultant for Sanofi, Olympus, Stryker and Optinose which are not affiliated with this manuscript. Zachary M. Soler is a consultant for Olympus, Optinose, Novartis, Regeneron, and Sinusonic which are not affiliated with this manuscript.
Footnotes
The abstract for this manuscript was accepted for podium presentation to the American Rhinologic Society, Summer Sinus Symposium, at Rhinoworld Chicago, Chicago, IL, June 5–9, 2019. Abstract confirmation #AD0NRB5XDN
REFERENCES
- 1.Smith KA, Orlandi RR, Rudmik L. Cost of adult chronic rhinosinusitis: A systematic review. Laryngoscope. 2015. doi: 10.1002/lary.25180. [DOI] [PubMed] [Google Scholar]
- 2.Soler ZM, Smith TL, Alt JA, Ramakrishnan VR, Mace JC, Schlosser RJ. Olfactory-specific quality of life outcomes after endoscopic sinus surgery. Int Forum Allergy Rhinol. 2016;6(4):407–413. doi: 10.1002/alr.21679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rudmik L Economics of Chronic Rhinosinusitis. Curr Allergy Asthma Rep. 2017;17(4). doi: 10.1007/s11882-017-0690-5. [DOI] [PubMed] [Google Scholar]
- 4.Hopkins C, Gillett S, Slack R, Lund VJ, Browne JP. Psychometric validity of the 22-item Sinonasal Outcome Test. Clin Otolaryngol. 2009;34(5):447–454. doi: 10.1111/j.1749-4486.2009.01995.x. [DOI] [PubMed] [Google Scholar]
- 5.Soler ZM, Kohli P, Storck KA, Schlosser RJ. Olfactory Impairment in Chronic Rhinosinusitis Using Threshold, Discrimination, and Identification Scores. Chem Senses. July 2016. doi: 10.1093/chemse/bjw080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.DeConde AS, Mace JC, Alt JA, Schlosser RJ, Smith TL, Soler ZM. Comparative effectiveness of medical and surgical therapy on olfaction in chronic rhinosinusitis: a prospective, multi-institutional study. Int Forum Allergy Rhinol. 2014;4(9):725–733. doi: 10.1002/alr.21350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Patel ZM, Thamboo A, Rudmik L, Nayak JV, Smith TL, Hwang PH. Surgical therapy vs continued medical therapy for medically refractory chronic rhinosinusitis: a systematic review and meta-analysis. Int Forum Allergy Rhinol. 2017;7(2):119–127. doi: 10.1002/alr.21872. [DOI] [PubMed] [Google Scholar]
- 8.Mattos JL, Schlosser RJ, DeConde AS, et al. Factor analysis of the questionnaire of olfactory disorders in patients with chronic rhinosinusitis. Int Forum Allergy Rhinol. 2018;8(7). doi: 10.1002/alr.22112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rudmik L, Hopkins C, Peters A, Smith TL, Schlosser RJ, Soler ZM. Patient-reported outcome measures for adult chronic rhinosinusitis: A systematic review and quality assessment. J Allergy Clin Immunol. 2015;136(6):1532–1540e2. doi: 10.1016/j.jaci.2015.10.012. [DOI] [PubMed] [Google Scholar]
- 10.Mattos JL, Schlosser RJ, Storck KA, Soler ZM. Understanding the relationship between olfactory-specific quality of life, objective olfactory loss, and patient factors in chronic rhinosinusitis. Int Forum Allergy Rhinol. May 2017. doi: 10.1002/alr.21940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sahlqvist S, Song Y, Bull F, Adams E, Preston J, Ogilvie D. Effect of questionnaire length, personalisation and reminder type on response rate to a complex postal survey: Randomised controlled trial. BMC Med Res Methodol. 2011. doi: 10.1186/1471-2288-11-62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harel D, Baron M, CSRG Investigators. Methods for shortening patient-reported outcome measures. Stat Methods Med Res. August 2018:962280218795187. doi: 10.1177/0962280218795187. [DOI] [PubMed] [Google Scholar]
- 13.Rosenfeld RM, Piccirillo JF, Chandrasekhar SS, et al. Clinical Practice Guideline (Update): Adult Sinusitis. Otolaryngol -- Head Neck Surg. 2015;152(2 Suppl):S1–S39. doi: 10.1177/0194599815572097. [DOI] [PubMed] [Google Scholar]
- 14.Simopoulos E, Katotomichelakis M, Gouveris H, Tripsianis G, Livaditis M, Danielides V. Olfaction-associated quality of life in chronic rhinosinusitis: adaptation and validation of an olfaction-specific questionnaire. Laryngoscope. 2012;122(7):1450–1454. doi: 10.1002/lary.23349. [DOI] [PubMed] [Google Scholar]
- 15.Walters SJ, Campbell MJ. The use of bootstrap methods for analysing health-related quality of life outcomes (particularly the SF-36). Health Qual Life Outcomes. 2004. doi: 10.1186/1477-7525-2-70. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.