|Home | About | Journals | Submit | Contact Us | Français|
We assessed cancer survivors' perceptions of the quality of their follow-up care.
We surveyed a population-based cohort of leukemia, bladder, and colorectal cancer survivors diagnosed 2 to 5 years previously in northern California (N = 623; participation rate, 69.2%; overall response rate, 49.2%). Data were collected between April 2003 and November 2004. Ten scales assessed survivors' perceptions of different aspects of care in the last 12 months, and an eleventh scale measured their overall ratings of care.
On nine of the 11 scales, mean scores ranged from 88 to 97 on a 0 to 100 response format, indicating very positive experiences. The two areas where quality perceptions were lower were discussions about health promotion and the physician's knowledge of the whole patient. In adjusted analyses, those without private health insurance (P = .02) and Hispanic and Asian survivors compared with whites (P < .001) reported worse timeliness of care. Survivors who had multiple comorbidities reported better scores on timeliness of care (P < .01) and physicians' knowledge (P = .05) than survivors without any comorbidity. Length of the patient-physician relationship was the variable most consistently found to be significantly associated with survivors' quality assessments. Physicians' information exchange had the strongest relationship with overall ratings of care, followed by physicians' affective behavior, their knowledge of the survivor, and survivors' perceptions of coordination of care (P < .001 for all).
Delivery of quality follow-up care to cancer survivors may require efforts to improve patient-centered communication and coordination. Special emphasis may need to be placed on health promotion discussions and adoption of a whole-person orientation.
Cancer survivors require substantial follow-up care to monitor for disease recurrence, to check for late effects of treatment, and to manage ongoing symptoms and adverse effects.1 Other goals of follow-up care include facilitating survivors' psychosocial adjustment to their illness and promoting healthy lifestyles.2,3 Although the importance of follow-up care has been recognized by clinicians, researchers, and policy makers,4,5 systematic assessments of the quality of care delivered to cancer survivors are limited.6
Existing studies have typically focused on understanding optimal models for delivering follow-up care and on developing and evaluating evidence-based guidelines for surveillance of cancer survivors.7–12 More recently, attention has been given to evaluating the role of care plans in facilitating the delivery of high-quality follow-up care.13 However, evaluation of the patient-centered aspects of care such as survivors' perspectives on access, communication with health care professionals, and perceptions of care coordination is virtually nonexistent.14 The few studies in oncology that have assessed patient perspectives have typically focused on the diagnosis and treatment phases of care.15–19 To provide a more comprehensive picture of the quality of follow-up care delivered to cancer survivors, systematic evaluations of the perspectives of survivors are needed to complement efforts that focus on more technical aspects of quality such as evaluation of the concordance of follow-up care practices with evidence-based guidelines.20–23
To better understand cancer survivors' experiences of receiving follow-up care, the National Cancer Institute sponsored the Assessment of Patient Experiences of Cancer Care (APECC) study, a population-based study of cancer survivors in northern California. This study had the following objectives: develop and test a comprehensive survey to assess cancer survivors' perceptions of the quality of their follow-up care; identify the sociodemographic, clinical, and follow-up care–related factors associated with survivors' quality-of-care assessments; and evaluate the association between survivors' assessments of the quality of individual aspects of care with their overall ratings of care.
We recruited cancer survivors from the Cancer Prevention Institute of California's cancer registry, a member of National Cancer Institute's Surveillance, Epidemiology, and End Results program. To be eligible, survivors had to read English; be diagnosed with leukemia or bladder or colorectal cancer between June 1, 1999, and May 31, 2001 (ie, 2 to 5 years before enrolling onto APECC); be at least 20 years old at diagnosis; have received cancer treatment; have the cancer of interest as their first cancer diagnosis; not have any other cancer between their initial diagnosis and the start of the study; and have no objections from their physician of record to their participation. We focused on leukemia and bladder and colorectal cancer because we wanted to test the survey measures across multiple cancers that were applicable to both sexes and individuals with a broad age range.
Figure 1 presents the study's recruitment flowchart. On the basis of information from the registry and physician consent, we identified 2,983 survivors to be potentially eligible for the study. We mailed them an initial letter describing the study, informing them about a $25 participation incentive, and indicating that an interviewer would call to screen them for eligibility. Of the 2,492 survivors who were located and screened, 1,572 (63.1%) were deemed eligible. Of these, 1,118 survivors agreed to participate and were mailed a self-reported survey. Surveys were returned by 774 survivors; the overall participation rate was 69.2% (774 of 1,118 survivors), and the overall response rate was 49.2% (774 of 1,572 survivors). Of the 774 survivors who provided data, 623 had received follow-up care in the last 12 months and responded to items assessing patients' care experiences. These 623 survivors form our analytic sample. All data collection procedures took place between April 2003 and November 2004. Study procedures were approved by the Institutional Review Board of the Cancer Prevention Institute of California.
We collected self-reported data on survivors' sociodemographic characteristics (age, sex, race/ethnicity, education, marital status, insurance status, and residence in a medically underserved area), their clinical characteristics (cancer type, perceived health status, number of comorbidities, years since diagnosis, type of treatments received, and whether or not their cancer was in remission), and several follow-up care–related variables (length of relationship with physician, number of visits in the last 12 months, physician's sex, survivor-physician sex match, physician specialty, and setting of care).
The APECC survey included 33 items assessing survivors' perceptions of the quality of their follow-up care in the last 12 months. Of these 33 items, 30 were divided into 10 conceptually distinct scales measuring survivors' experiences in the following five broad areas: access to care (six items, three scales: getting needed care, timeliness of care, and waiting time in physicians' office); interaction with physicians (17 items, three scales: information exchange, affective behavior, and physicians' knowledge); interaction with other members of the health care team (four items, two scales: interaction with nurses and interaction with office staff); discussion of health promotion (two items, one scale); and perceptions of coordination of care (one item). The remaining three items (one scale) assessed survivors' overall ratings of their care (Table 2 provides the exact wording of all items).
Where possible, items were adapted from existing surveys including the Consumer Assessment of Healthcare Providers and Systems surveys24–27 (11 items), the Primary Care Assessment Survey28–30 (four items), and the Ambulatory Care Experiences Survey31,32 (five items). Another two items were adapted from a study of experiences of patients with colorectal cancer.15 In addition, 11 new items were developed by the APECC investigators (noted in Table 2). Although the wording of existing items was modified for relevance to the context of follow-up care, we preserved their original response formats. A majority of the items had one of the following four response options: not a problem, a small problem, or a big problem; never, sometimes, usually, or always; yes definitely, yes somewhat, or no; or poor, fair, good, very good, or excellent. As suggested in other studies of patient experiences of care,15,16 for descriptive purposes only, for each item, we considered a response of less than the most positive option (ie, a response other than not a problem, always, yes definitely, or excellent) to be an indicator of suboptimal quality. All items underwent cognitive testing with nine cancer survivors (with diversity in age, race, sex, and cancer type) to ensure that the questions and response options were understandable and related to the concept being measured.
To confirm the 10-factor/scale structure of the APECC items assessing patient perspectives on individual aspects of follow-up care, we conducted a confirmatory factor analysis using Mplus version 4.21 (Muthen & Muthen, Los Angeles, CA). We evaluated the statistical fit of the 10-factor model to the data using multiple fit indices, as follows: comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). Generally, CFI values greater than 0.95, RMSEA values less than 0.06, and SRMR values less than 0.09 are considered to indicate good model fit.33 The confirmatory factor analysis indicated a reasonably good fit for the 10-factor model (CFI = 0.93, RMSEA = 0.04, and SRMR = 0.04). Scores for all of these 10 variables and the overall ratings of care variable were created by computing the mean of the scores on the individual items comprising each variable. All scores were then linearly transformed to a 0 to 100 range, with a higher score representing more positive quality assessments. All but two of the 10 multi-item scales had acceptable internal consistency reliability (Cronbach's α > .75; details on score distribution are presented in Table 3 and discussed later).
We examined the associations of various sociodemographic, clinical, and follow-up care–related variables with survivors' quality assessments by estimating analysis of covariance models. Patient sex, residence in a medically underserved area, remission status, and time since diagnosis were not associated in bivariate analyses with any of the variables assessing survivors' quality perceptions (P > .1) and were not considered for adjusted analyses. We did not include physician specialty and treatment in our models because they were highly correlated with cancer type (P < .001). Among the bladder cancer survivors, 91% saw a urologist; all but two leukemia survivors saw a hematologist/oncologist; and only colorectal cancer survivors saw physicians of different specialties (primary care physician, 15%; oncologist, 62%; other specialist, 23%). Similarly, a majority of bladder cancer survivors had surgery only (74%), 89% of leukemia survivors had chemotherapy with or without radiation, and 66% of colorectal cancer survivors had surgery and adjuvant therapy (chemotherapy, radiation, or both). We conducted linear regression analysis to examine the associations between the 10 variables assessing cancer survivors' perceptions of quality of individual aspects of care and their overall care ratings. We controlled for all the sociodemographic, clinical, and follow-up care–related variables analyzed in the analysis of covariance models. All models were estimated using SPSS version 14.0 (SPSS, Chicago, IL). P ≤ .05 was considered to be statistically significant.
Table 1 lists the various characteristics of our sample. Twenty-six percent of participants were bladder cancer survivors, 60% were colorectal cancer survivors, and 14% were leukemia survivors. Approximately half of survivors had more than one comorbidity, and 80% received care from the same physician for more than 2 years. Compared with the eligible nonrespondents, respondents were relatively younger (P < .001) and were less likely to be bladder cancer survivors and more likely to be survivors of colorectal cancer or leukemia (P < .001). There were no differences between respondents and nonrespondents regarding sex, race/ethnicity, or year of diagnosis (data not shown).
Table 2 lists the percentage of survivors who reported less than optimal quality in response to the 33 APECC items. Reports of suboptimal quality ranged from 4% to 77%. Although only 4% of survivors reported problems with access to specialists, 33% reported that their physician did not always encourage them to ask questions, 59% reported limited or no discussion on health promotion and prevention topics, and 77% indicated that their physician had less than excellent knowledge of them as a person.
Table 3 lists the score distribution for the 10 variables assessing survivors' perceptions of quality of individual aspects of care as well as their overall care ratings. With the exception of the health promotion and physicians' knowledge scales, all scores indicated very positive care experiences.
As shown in Table 4, older survivors, survivors who were married or living as married, survivors who perceived themselves to be in better health, and survivors who had been seeing the same physician for more than 2 years had significantly more positive perceptions of quality on multiple aspects of care. Survivors belonging to minority racial/ethnic subgroups (P < .001) and those without private health insurance (P = .02) reported worse experiences on timeliness of care. Survivors who received care in a health maintenance organization clinic reported less problems with timeliness of care (P < .05) and waiting time (P < .001). With the exception of health promotion, survivors' perceptions of quality of care did not vary with type of cancer. Survivors who had multiple comorbidities reported better timeliness of care (P < .01) and greater physicians' knowledge (P = .05) than those who did not have any comorbidity.
As shown in Table 5, survivors' assessments of the quality of interactions with their physician (P < .001), nurses (P = .05), and office staff (P = .001), as well their perceptions of care coordination (P < .001), were significantly associated with their overall ratings of care. The strongest association with overall ratings was observed for information exchange. With the exception of getting needed care, access-related variables were not significantly associated with overall ratings; discussion of health promotion was also not significantly associated with overall ratings.
Cancer survivors in our study had positive quality perceptions for many aspects of their follow-up care. However, multiple areas for improvement were also identified. The top five items where survivors reported suboptimal quality were related to physician knowledge and health promotion. More than 60% of survivors reported that their physician lacked complete knowledge of how their quality of life had been affected by their cancer and its treatment, and more than 75% of survivors perceived their physician to have a less than ideal understanding of them as a person. To facilitate optimal patient outcomes, the Institute of Medicine's Cancer Care for the Whole Patient: Meeting Psychosocial Health Needs report34 calls for clinicians to adopt a whole-person orientation while interacting with patients. Survivors in our study perceived this to be an area for improvement. More than 50% of survivors also reported suboptimal discussion of things they could do to maintain healthy lifestyles and prevent future illnesses. Although disease prevention and enhancement of well-being have been identified as essential components of survivorship care,4 our data suggest that communication about health promotion during follow-up care visits is limited.
There were several other areas of follow-up care where at least one in four survivors reported suboptimal quality that could be targets for improvement. These included delays in start of appointments, perceived lack of care coordination among the care team, less than ideal help provided by the office staff, and experiences of suboptimal information exchange with the physician. These reports suggest that to deliver optimal follow-up care, quality improvement efforts would need to focus not only on behaviors of individual physicians and health care team members, but also on redesigning workflow in the office practice to improve efficiency, communication, and coordination.
Consistent with existing studies, older survivors reported more positive experiences, whereas survivors who perceived their health to be poorer reported worse experiences.35 Contrary to our expectations, survivors with more comorbidities reported better timeliness of care and physician knowledge. It is possible that cancer survivors who live with multiple comorbidities have more intense interactions with the health care system, making them more experienced consumers. Support for this assertion is also obtained from our finding that survivors who had more contact with their providers in terms of number of visits also reported more positive experiences on timeliness of care and physician knowledge. It is also possible that given busy practices, clinicians tend to selectively pay more attention to their patients who have a greater cumulative disease burden. More studies are needed to disentangle the reasons for the intriguing association between disease burden and patient perceptions of quality.
Consistent with existing studies, Hispanic and non-Hispanic Asian survivors reported worse scores on timeliness of care than non-Hispanic white survivors.15,36–38 Given that survivors who did not have private health insurance also reported worse experiences with obtaining timely care suggests the potential for disparities in access to follow-up care between potentially underserved populations of survivors, such as minorities, the uninsured, and the underinsured, and the rest of survivors. Post hoc analysis exploring the interaction effect of race/ethnicity and insurance status on timeliness of care revealed a nonsignificant interaction, suggesting that being of a minority race/ethnicity and being uninsured or underinsured may be independent risk factors for not receiving timely care. The impact of not receiving timely follow-up care on patient outcomes needs to be explored.
Survivors who had been receiving care from the same physician for more than 2 years reported more positive experiences on all but two quality indicators compared with survivors who knew their physician for less than a year. A longer tenure of the patient-physician relationship may result in better understanding of patient needs by the physician, thus resulting in more positive experiences.39 It is also possible that more positive evaluations among survivors who knew their physician for a longer time are a reflection of a halo effect where survivors are more likely to ignore lapses in quality from a physician with whom they have established a long-term relationship than from one who they do not know well. Length of the patient-physician relationship and quality-of-care perceptions are likely to mutually influence each other, and their relationship can be better teased out within the context of longitudinal study designs.
Perceived quality of information exchange between physicians and survivors had the strongest relationship with overall ratings of care, followed by physicians' affective behavior, physicians' knowledge of the survivor, and survivors' perceptions of coordination of care. These associations further highlight the importance of optimizing the various functions of patient-physician communication as well as communication among the various members of the health care team.40 Interestingly, although survivors reported limited communication related to health promotion, such lack of communication did not have any impact on their overall ratings. These findings suggest that although clinicians may not be devoting adequate time during follow-up care visits to facilitating healthy lifestyles among survivors, cancer survivors themselves may not be expecting such discussions. Whether survivors' lack of expectation about health promotion discussions might be driven by a lack of awareness of the importance of healthy behaviors to their overall health or is a result of their health promotion needs being met by sources other than their follow-up care physician is not clear and needs further exploration.
Our findings should be interpreted in the light of potential limitations. The most important limitation is the cross-sectional nature of the study. Causal implications cannot be inferred from the significant associations reported in this study; findings need to be replicated within the context of longitudinal study designs. We recruited participants from a population-based sample of leukemia, bladder, and colorectal cancer survivors in Northern California. Our findings may not be generalizable to survivors of other common cancers such as breast and prostate cancer or to survivors in other parts of the United States. Further validation of the APECC measures in other cancer survivor populations is needed. Although we were able to compare the experiences of Hispanic and non-Hispanic Asian survivors with those of non-Hispanic white survivors, we could not do the same for non-Hispanic African American survivors because of the small size of the African American subgroup in our sample.
Despite potential limitations, the APECC study provides unique data on the follow-up care experiences of cancer survivors. With the aging of the population and the dissemination of effective therapeutic interventions in oncology, the ranks of cancer survivors are likely to increase substantially in the future. Efforts at optimizing quality of care delivered to these survivors in the future should focus not only on the technical aspects but also on the patient-centered aspects of care. The APECC study lays the foundation for systematic evaluations of patient-centeredness of follow-up care provided to cancer survivors in the United States.
Support for data collection was provided by the National Cancer Institute (Contract No. N01-PC-35136) as a contract to the Cancer Prevention Institute of California (formerly known as the Northern California Cancer Center). R.D.H. was also supported in part by the University of California, Los Angeles (UCLA) Resource Center for Minority Aging Research/Center for Health Improvement in Minority Elderly, National Institutes of Health (NIH)/National Institute on Aging (NIA) Grant No. P30-AG021684, the UCLA/Drew Project Centers of Excellence in Partnerships for Community Outreach and Research on Disparities in Health and Training, National Center on Minority Health and Health Disparities Grant No. 2P20MD000182, and the UCLA Older Americans Independence Center (NIH/NIA Grant No. P30-AG028748).
This article reflects the personal opinions of N.K.A. and S.B.C. and does not convey any official position of the National Cancer Institute.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
The author(s) indicated no potential conflicts of interest.
Conception and design: Neeraj K. Arora, Bryce B. Reeve, Ingrid Oakley-Girvan
Financial support: Steven B. Clauser
Provision of study materials or patients: Ingrid Oakley-Girvan
Collection and assembly of data: Ingrid Oakley-Girvan
Data analysis and interpretation: Neeraj K. Arora, Bryce B. Reeve, Ron D. Hays, Steven B. Clauser
Manuscript writing: Neeraj K. Arora, Bryce B. Reeve, Ron D. Hays, Steven B. Clauser, Ingrid Oakley-Girvan
Final approval of manuscript: Neeraj K. Arora, Bryce B. Reeve, Ron D. Hays, Steven B. Clauser, Ingrid Oakley-Girvan