International Journal for Quality in Health Care Advance Access originally published online on June 29, 2006
International Journal for Quality in Health Care 2006 18(4):266-274; doi:10.1093/intqhc/mzl014
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Factors that influence cancer patients’ overall perceptions of the quality of care
1 Department of Health Policy, Management, and Evaluation, University of Toronto, 2 Cancer Care Ontario, and 3 Faculty of Nursing, University of Toronto, Toronto, Ontario, Canada
Objective. This study outlines predictors of cancer patients’ overall perceptions of the quality of care.
Design and setting. Our sample included 2790 patients who received cancer care services during 2004 in 15 comprehensive cancer programmes across Ontario, Canada. Patients were classified into three groups: those receiving both chemotherapy and radiotherapy (n = 752), those receiving only chemotherapy (n = 1044), and those receiving only radiotherapy (n = 994). An ordinal logistic regression model for each patient group was performed to determine which variables most affected the probabilities of the patients’ overall evaluations of the quality of care. Potential control variables were patients’ age, sex, type of cancer, self-assessed health, and who completed the survey.
Results. Among seven common predictors of the overall quality perception across the three models, four should be of particular interest because patients perceived them as relatively problematic aspects of care. These are ‘was informed about follow-up care after completing treatment’, ‘knew next step in care’, ‘knew who to go to with questions’, and ‘providers were aware of test results’. These predictors explained between 25 and 34% of the variance (depending on the model) of the overall perception of quality. The explanatory power of these predictors did not change across sex and age group. ‘Self-assessed health’ was the only control variable that remained in all three models.
Conclusions. From a practical perspective, improvement efforts are best focused on factors that are strong predictors as well as on those for which there is a low score. Thus, on the basis of this study, practitioners’ improvement efforts might be constructively focused on the four predictors mentioned above.
Keywords: cancer patients, patient satisfaction, perceptions of the quality of care, predictors
Address reprint requests to Adalsteinn D. Brown, Department of Health Policy, Management, and Evaluation, University of Toronto, 155 College Street, Room 425, Toronto, Ontario, Canada M5T 3M6. E-mail: adalsteinn.brown{at}utoronto.ca
Accepted for publication April 28, 2006.
Assessment of patient satisfaction with cancer care experiences has increasingly gained attention over the last decade. To date, considerable interest has been placed in evaluating care experiences with treatment [1–2345678910], detection and diagnosis [6–78,11–12131415], and follow-up care [3,11–2113,16–1718]. These service experiences form part of the cancer care continuum which involves risk assessment, primary prevention, detection and diagnosis, treatment, recurrence surveillance, and end-of-life care [19]. A few studies have focused on cancer patients’ care experiences with end-of-life care [9,20,21], and it appears that patient satisfaction with risk assessment or primary prevention has not been investigated. The most common attributes or dimensions evaluated in the current cancer literature are overall satisfaction, level of information received, and patient–provider interpersonal interactions, whereas among the least common are symptom and pain management, waiting times, and coordination of care.
Some studies found information received, technical competences, interpersonal and communication skills, time spent talking with doctors and nurses, accessibility and coordination of care, waiting times, and patients’ emotional needs as important or priority areas to improve cancer care services [2,4,13]. Other studies, and using different approaches, identified those factors that best predicted a variable that measured cancer patients’ overall assessments of the care received [1,5,9,12,15]. Among them were aspects related to doctors, nurses, information received, and time issues. Most of these studies used small sample sizes, were conducted in one facility, and evaluated patient perceptions of the service provided in one phase of the cancer care continuum.
Using the results of a Picker-based suite of surveys, this study aimed at identifying those factors that most influenced a variable that measured cancer patients’ overall perceptions of the quality of care received. The multifacility sample of 2790 cancer patients used in this analysis was representative of close to 70 000 patients who received care services across Ontario, the most populated province in Canada, during 2004. Using ordinal logistic regression modelling, we identified predictors for three groups of patients: those receiving both chemotherapy and radiotherapy, those receiving only chemotherapy, and those receiving only radiotherapy. This analysis might ultimately help practitioners focus their efforts when planning initiatives aimed at improving the patient care experience and is in alignment with some of the six aims for improvement identified by the Institute of Medicine, those referring to ‘patient-centered’ and ‘efficient’ in particular [22].
 |
Patients and methods |
|---|
 Top Patients and methods ResultsDiscussionReferences |
|---|
Cancer Care Ontario—the provincial government’s principal adviser on cancer issues—organized the Ambulatory Oncology Patient Satisfaction Survey (AOPSS) for patients who had received cancer care services or treatment during 2004 across the province of Ontario, Canada. The instrument—based on the Picker suite of surveys—was developed and validated in Canada by the National Research Corporation who surveyed close to 5000 cancer patients from three provinces during the winter of 2003 [23]. The AOPSS includes 79 closed-ended items plus one open-ended item. The aspects of care evaluated in this survey cover mainly two phases of the cancer care continuum—detection and diagnosis and treatment—and relate to cancer diagnosis, treatment planning, tests, surgery, chemotherapy, radiotherapy, symptom management, health care providers, and overall impressions. The survey results were collected by a third party, and patients received a cover letter describing the survey purpose, outlining that completion of the survey was voluntary, and assuring anonymity.
During September and October 2004, the AOPSS was mailed using a two-wave mail methodology to 8521 cancer patients who had visited 15 comprehensive cancer care programmes between March and August 2004. These patients were randomly sampled from a database of close to 70 000 patients who were the actual number of patients served by the 15 programmes between March and August 2004. The purpose of the AOPSS was to capture the experiences from patients who were receiving cancer care services within the last 6 months. The first question instructed the patient to return the survey if he/she had not received care services or had only received follow-up care in the last 6 months. A total of 5015 surveys were returned between September and December 2004 (overall response rate: 58.7%); however, a total of 1569 patients answered only the first question and returned the survey. To have valid regression models, we divided the patients into three major groups: those that only received chemotherapy (1044), those that only received radiotherapy (994), and those that received both chemotherapy and radiotherapy (752). Sixty-five per cent of these patients were first told about their cancer diagnoses less than a year ago, whereas 28% between 1 and 5 years ago and 7% >5 years ago. Sixty per cent of these patients also underwent surgery. Having included all cases in one regression model would have resulted in many missing cases because those receiving only chemotherapy would not have answered the radiotherapy section and vice versa. Patients who did not declare what kind of cancer care service they received (327) and those who underwent only surgery (329) were not included in this analysis. The latter group did not answer the sections about chemotherapy and radiotherapy; therefore, having included them would have increased again the number of missing cases for the regression purposes. Table 1 summarizes relevant information of the three major groups analysed in this study (total N = 2790). The groups statistically differed on sex, age, and type of cancer.
|
Similar to other studies [9,12,15], we selected a single item from the survey as the dependent variable, that is ‘overall, how would you rate the quality of all of your care in the past 6 months?’. It has also been argued that this approach is more practical than the use of longer or multi-item measures [15]. The way this item was worded (‘all of your care in the past 6 months’) limited the use of the independent variables to those aspects from the questionnaire that described situations within the last 6 months. For example, some patients were first told about their cancer diagnoses between 2 and 5 years ago, and therefore, items such as ‘were you told of your diagnosis in a sensitive manner?’ refer to that period. Because we are trying to predict patients’ perceptions of all care received in the last 6 months (as the dependent variable states), it appeared reasonable not to include aspects of care referring to those patient–provider encounters that occurred >6 months ago. Therefore, all items related to diagnosis, treatment planning, tests, and surgery were discarded as independent variables because those services may have not been provided in the past 6 months. We also discarded a few questions that were too broad or those in which it was difficult to conceptualize a specific aspect of care (e.g. ‘did a care provider go out of his or her way to help you or make you feel better?’). Finally, we excluded all survey questions with >40% of responses missing. As a result, 29 items were selected as independent variables. They related to chemotherapy, radiotherapy, symptom management, and health care providers. All items were statistically associated with the dependent variable (Spearman’s r range: 0.18–0.42, P < 0.0001). This single-item approach to the independent variables is similar to other studies [9,12,15] and appears to be more practical when planning improvement initiatives. In addition, the majority of the inter-item correlations were in the range 0.10–0.40. This was also important to demonstrate a reasonable degree of item independence. Table 2 presents all items used in this analysis, their scales, and their ‘problem scores’ for each of the three patient groups. A ‘problem score’ is the percentage of patients answering all but not the best evaluation possible. The higher the percentage, the more patients perceive a determined aspect of care as a problem. In other words, a problem is defined as an aspect of health care that could, in the eyes of the patient, be improved upon [24].
|
We also included five potential control variables: sex, age, type of cancer, self-assessed health, and who completed the survey. Using the stepwise selection procedure, we entered the independent variables into an ordinal logistic regression model to determine which variables most affected the probabilities of the patients’ overall evaluations of the quality of care (the dependent variable, a 5-point scale). In addition, we tested the interaction effects of sex and age group with all the aspects of care. We performed three separate models: one for patients receiving both chemotherapy and radiotherapy (n = 752), one for patients receiving only chemotherapy (n = 1044), and one for patients receiving only radiotherapy (n = 994). For every independent variable in the final models, an odds ratio was estimated. It indicates the number of times the odds of rating the overall quality of care as ‘excellent’ versus ‘not excellent’ (i.e. ‘very good’, ‘good’, ‘fair’, or ‘poor’) increase for every 1-point increase on the scale of the independent variable. Because the response variable has five ordered categories, the interpretation above is also true for odds of (i) ‘excellent’ or ‘very good’ versus ‘good’, ‘fair’, or ‘poor’; (ii) ‘excellent’, ‘very good’, or ‘good’ versus ‘fair’ or ‘poor’; and (iii) ‘excellent’, ‘very good’, ‘good’, or ‘fair’ versus ‘poor’. All independent variables were coded in the same direction to facilitate interpretation. Because all data used in this analysis were ordinal or nominal, the non-parametric chi-square test was performed in all comparisons. Before performing the regressions, missing values were replaced using the Gaussian model in S-Plus (Insightful Corporation, Seattle, WA, USA). All analyses were performed using the SAS System for Windows (SAS Institute, Cary, NC, USA).
|
Results |
|---|
|
TopPatients and methodsResultsDiscussionReferences |
|---|
The 1569 patients who returned their surveys with only question one answered were considered non-respondents for the purpose of respondent/non-respondent comparisons (3446/5075). Although we found significant differences (P < 0.001) between respondents and non-respondents on gender and age group, they were not meaningful as per Cohen’s definition of effect sizes related to chi-square (small = 10%, moderate = 30%, and large = 50%) [25]. Respondents were slightly older and had a larger proportion of males than non-respondents. A total of 50.5% of non-respondents aged 65 and over, whereas 51.5% of respondents aged 65 and over (effect: 1% difference). Non-respondents aged 55–64 were 22.7%, and respondents in the same category were 24.6% (effect: 1.9% difference). Male non-respondents were 41.0%, whereas male respondents were 44.9% (effect: 3.9% difference).
Patients’ perceptions of their health care experience
Table 2 summarizes the problem scores for all items across the three patient groups. Radiotherapy patients tended to report few problems with the aspects of care than did chemotherapy and chemoradiotherapy patients. The least problematic aspects in all three patient groups were ‘treated with respect and dignity by care providers’ and ‘could trust care providers with confidential info’. The most problematic aspects were ‘enough info on relationship changes’ and ‘waited longer than expected for chemo treatment’. Table 3 reports patients’ overall perceptions of the quality of care across respondent characteristics for each of the three patient groups. Male, older, and healthier patients tended to rate the overall quality of care higher than female, younger, and less healthy patients. Patients with prostate cancer tended to rate the overall quality of care higher than those with other cancer types. Finally, patients’ overall perceptions of the quality of care were also lower in cases where someone other than the patient completed the survey.
|
Predictors of patients’ overall perceptions of the quality of care
All final three logistic regression models are presented in Tables 4, 5, and 6, and all the variables are statistically significant at P < 0.05 level. The model for chemoradiotherapy patients explained 37% of the variance of the overall quality perception. The model for chemotherapy patients explained 45% of the variance, and the model for radiotherapy patients explained 38% of the variance. ‘Self-assessed health’ was the only control variable that remained in all three models. The interaction effect analyses revealed that none of the predictors in all three models affected the dependent variable differently across gender and age group. The largest odds ratio was observed in the model involving radiotherapy patients for the independent variable ‘treated with respect and dignity by care providers’ (odds ratio: 4.38). One interpretation of this value is that if the scale of ‘treated with respect and dignity by care providers’ (a 3-point scale) is increased one point, the odds of rating the overall quality of care by the patient as ‘excellent’ versus ‘not excellent’ (i.e. ‘very good’, ‘good’, ‘fair’, or ‘poor’) increase 4.38 times.
|
|
|
|
Discussion |
|---|
|
TopPatients and methodsResultsDiscussionReferences |
|---|
This study identified predictors of a variable that measured patients’ overall perceptions of the quality of care received. Three patient groups were analysed: chemoradiotherapy patients (n = 752), chemotherapy patients (n = 1044), and radiotherapy patients (n = 994). The variance of the overall quality perception explained in each model is somewhat similar to the models used by Skarstein et al. [5] and Walker et al. [15] which explained 35 and 37%, respectively. This distinct segregation of the data might help health care managers and providers be more oriented when planning strategies aimed at improving the patient care experience. Such strategies may become more significant because evidence suggests that satisfied patients are more likely to cooperate with their treatment [1,2,5,9,15,26] which is, in turn, associated with better clinical outcomes [15].
Among seven common predictors of the overall quality perception across the three models, four are of particular interest because patients perceived them as relatively problematic aspects of care. These are ‘was informed about follow-up care after completing treatment’, ‘knew next step in care’, ‘knew who to go to with questions’, and ‘providers were aware of test results’. Together these explained between 25 and 34% of the variance (depending on the model) of the overall quality perception. These predictors are important from a practical perspective because improvement initiatives should be placed in aspects that score relatively low but are highly associated with the dependent variable [2,13].
For the model involving chemoradiotherapy patients, the two aspects of care pereived as most problematic were ‘waited longer than expected for chemo treatment’ and ‘enough info on relationship changes’. These predictors explained together 13% of the variance of the overall quality perception (37%, the final model). From the survey responses, we estimated that this patient group waited on average 56 minutes between the scheduled appointment and the chemotherapy treatment. Therefore, if one would like to positively affect patients’ overall perceptions of the quality of care, efforts should be directed to reducing this waiting time. However, if this is not feasible, an attempt should be made to make delays feel shorter than they actually are, through control of environmental factors [22]. For the model involving chemotherapy patients, the predictor ‘waited longer than expected for chemo treatment’ appeared again as the most problematic one. In addition, the predictors ‘providers were aware of medical history’ and ‘knew next step in care’ were aspects of care that were most problematic. Finally, for the model involving radiotherapy patients, the predictors ‘enough info on energy changes’ and ‘knew who to go to with questions’ appeared as the two most problematic aspects of care as perceived by patients. The interaction effect analyses also indicated that there is no evidence to embark on different improvement actions across gender and age group. This finding is similar to Walker et al. [15] who found no differences in predictors of patient satisfaction across gender and age group.
From the original five potential control variables, ‘self-assessed health’ remained in all three models revealing a positive association with the dependent variable. That is, the better the health status of a patient, the greater the likelihood of higher ratings on overall quality of care. This relationship is similar to other studies that have shown that health status influences patient satisfaction with care [26–272829]. In cancer settings, however, it appears that this relationship has not been extensively studied. Although we found considerable literature evaluating cancer care experiences [1–23456789101112131415161718,20,21], the instruments as well as the methodology used varied widely. For example, different from our study, other studies have included as predictor variables physiological aspects of the illness [5,15] as well as the severity of cancer [5]. The presence of coexisting or additional diseases (comorbidity) might also be related to patient satisfaction and could be included in future researches. Similarly, most studies have evaluated care experiences in one phase of care of the cancer continuum, and no study appears to have evaluated patient satisfaction across the full spectrum of cancer care from prevention to end-of-life care. This also suggests avenues for further research.
The findings of this study are limited to the instrument used, notwithstanding that the variations of this tool are widely used in ambulatory oncology settings and useful for most cancer programmes. For example, the survey uses items such as ‘how often were your care providers aware of your test results?’. This is different from most instruments which specifically evaluate various aspects of doctors, nurses, and other care providers such as technical competences and interpersonal and communication skills. Despite this limitation, we were able to compare our findings with those of others because the aspects of care that emerged from our three final models were mainly related to two main dimensions of care: information provided and waiting times. Skarstein et al. [5] and Fossa et al. [12] found aspects related to information received as significant predictors of overall satisfaction. Gourdji et al. [2] and Gesell and Gregory [13] found waiting time as an aspect of care that need to be prioritized. The results of our study are, thus, not surprising in the light of others who have identified similar aspects of care as relevant to patient satisfaction.
Although the low valid response rate in this study (40%) is a potential limitation, this might not necessarily jeopardize the generalizability of our results because the differences between respondents and non-respondents on age and gender were smaller than Cohen’s definition of a small effect size [25]. We also found seven cases in which the residuals were located >3 SD from the residual mean. We ran the models again, but the slight changes observed on the odds ratios were not enough to change our results. Finally, because patient satisfaction is considered one of multiple aspects in measuring the quality of care [26,29,30], our recommendations should be viewed only as one of many areas that need to be addressed in providing comprehensive access to high-quality health services at a reasonable cost—the major concern of developed countries [31].
|
References |
|---|
|
TopPatients and methodsResultsDiscussionReferences |
|---|
- Kavadas V, Barham CP, Finch-Jones MD et al. Assessment of satisfaction with care after inpatient treatment for oesophageal and gastric cancer. Br J Surg 2004; 91: 719–723.[CrossRef][Web of Science][Medline]
- Gourdji I, McVey L, Loiselle C. Patients’ satisfaction and importance ratings of quality in an outpatient oncology center. J Nurs Care Qual 2003; 18: 43–55.[Web of Science][Medline]
- Liberati A, Confalonieri C, Martino G et al. Patients’ assessment of quality of care: a survey of a group of breast cancer patients in Italy. Tumori 1985; 71: 491–497.[Web of Science][Medline]
- Wiggers JH, Donovan KO, Redman S et al. Cancer patient satisfaction with care. Cancer 1990; 66: 610–616.[CrossRef][Web of Science][Medline]
- Skarstein J, Dahl AA, Laading J et al. ‘Patient satisfaction’ in hospitalized cancer patients. Acta Oncol 2002; 41: 639–645.[CrossRef][Web of Science][Medline]
- Veronesi U, von Kleist S, Redmond K et al. Caring About Women and Cancer (CAWAC): a European survey of the perspectives and experiences of women with female cancers. Eur J Cancer 1999; 35: 1667–1675.[CrossRef][Web of Science][Medline]
- McPhail G, Wilson S. Women’s experience of breast conserving treatment for breast cancer. Eur J Cancer Care (Engl) 2000; 9: 144–150.
- Jansen SJT, Otten W, van de Velde CJH et al. The impact of the perception of treatment choice on satisfaction with treatment, experienced chemotherapy burden and current quality of life. Br J Cancer 2004; 91: 56–61.[CrossRef][Web of Science][Medline]
- Blanchard CG, Labrecque MS, Ruckdeschel JC et al. Physician behaviors, patient perceptions, and patient characteristics as predictors of satisfaction of hospitalized adult cancer patients. Cancer 1990; 65: 186–192.[CrossRef][Web of Science][Medline]
- Brédart A, Robertson C, Razavi D et al. Patients’ satisfaction ratings and their desire for care improvement across oncology settings from France, Italy, Poland, and Sweden. Psychooncology 2003; 12: 68–77.[CrossRef][Medline]
- Shilling V, Jenkins V, Fallowfield L. Factors affecting patient and clinician satisfaction with the clinical consultation: can communication skills training for clinicians improve satisfaction? Psychooncology 2003; 12: 599–611.[CrossRef][Medline]
- Fossa SD, Hjermstad MJ, Mork IH et al. Does the service at a large oncologic out-patient clinic satisfy the patients’ perceived need? Int J Health Care Qual Assur 1996; 9: 24–29.[Medline]
- Gesell SB, Gregory N. Identifying priority actions for improving patient satisfaction with outpatient cancer care. J Nurs Care Qual 2004; 19: 226–233.[Web of Science][Medline]
- Rehnberg G, Absetz P, Aro AR. Women’s satisfaction with information at breast biopsy in breast cancer screening. Patient Educ Couns 2001; 42: 1–8.[CrossRef][Web of Science][Medline]
- Walker MS, Ristvedt SL, Haughey BH. Patient care in multidisciplinary cancer clinics: does attention to psychosocial needs predict patient satisfaction? Psychooncology 2003; 12: 291–300.[CrossRef][Medline]
- Grunfeld E, Fitzpatrick R, Mant D et al. Comparison of breast cancer patient satisfaction with follow-up in primary care versus specialist care: results from a randomized controlled trial. Br J Gen Pract 1999; 49: 705–710.[Web of Science][Medline]
- Koinberg I, Holmberg L, Fridlund B. Satisfaction with routine follow-up visits to the physician – the needs of patients with breast cancer. Acta Oncol 2001; 40: 454–459.[Web of Science][Medline]
- Adlard JW, Joseph J, Brammer CV et al. Open access follow-up for lung cancer: patient and staff satisfaction. Clin Oncol 2001; 13: 404–408.
- Zapka JG, Taplin SH, Solberg LI et al. A framework for improving the quality of cancer care: the case of breast and cervical cancer screening. Cancer Epidemiol Biomarkers Prev 2003; 12: 4–13.
[Abstract/Free Full Text] - Twycross R, Harcourt J, Bergl S. A survey of pain in patients with advanced cancer. J Pain Symptom Manage 1996; 12: 273–282.[CrossRef][Web of Science][Medline]
- Bostrom B, Sandh M, Lundberg D et al. Cancer patients’ experiences of care related to pain management before and after palliative care referral. Eur J Cancer Care (Engl) 2004; 13: 238–245.
- Leddy KM, Kaldenberg DO, Becker BW. Timeliness in ambulatory care treatment. An examination of patient satisfaction and wait times in medical practices and outpatient test and treatment facilities. J Ambul Care Manage 2003; 26: 138–149.[Medline]
- National Research Corporation. Development and Validation of the Picker Ambulatory Oncology Survey Instrument in Canada. Lincoln, NE: 2003. Available online at http://www.cancercare.on.ca/qualityindex2006/outcomes/satisfaction/index.html.
- Jenkinson C, Coulter A, Bruster S. The Picker Patient Experience Questionnaire: development and validation using data from in-patient surveys in five countries. Int J Qual Health Care 2002; 14: 353–358.
[Abstract/Free Full Text] - Munro BH. Statistical Methods for Health Care Research. Philadelphia: Lippincott Williams & Wilkins, 2005.
- Pascoe GC. Patient satisfaction in primary health care: a literature review and analysis. Eval Program Plann 1983; 6: 185–210.[CrossRef][Medline]
- Westaway MS, Rheeder P, Van Zyl DG et al. Interpersonal and organizational dimensions of patient satisfaction: the moderating effects of health status. Int J Qual Health Care 2003; 15: 337–344.
[Abstract/Free Full Text] - Hall JA, Milburn MA, Roter DL et al. Why are sicker patients less satisfied with their medical care? Tests of two explanatory models. Health Psychol 1998; 17: 70–75.[CrossRef][Web of Science][Medline]
- Covinsky KE, Rosenthal GE, Chren MM et al. The relation between health status changes and patient satisfaction in older hospitalized medical patients. J Gen Intern Med 1998; 13: 223–229.[CrossRef][Web of Science][Medline]
- Donabedian A. The quality of care. How can it be assessed? JAMA 1998; 260: 1743–1748.
- Sullivan T, Evans W, Angus H et al. Strengthening the Quality of Cancer Services in Ontario. Ottawa: CHA Press, 2003.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Wessels, A. de Graeff, K. Wynia, H. J. Sixma, M. de Heus, M. Schipper, G. T. G. J. Woltjer, S. C. C. M. Teunissen, and E. E. Voest Medical oncology patients' preferences with regard to health care: development of a patient-driven questionnaire Ann. Onc., October 1, 2009; 20(10): 1708 - 1713. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Teno, J. C. Lima, and K. D. Lyons Cancer Patient Assessment and Reports of Excellence: Reliability and Validity of Advanced Cancer Patient Perceptions of the Quality of Care J. Clin. Oncol., April 1, 2009; 27(10): 1621 - 1626. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||