Critical Choices: What Information Do Patients Want When Selecting a Hospital for Cancer Surgery?

Abstract

Purpose:

Access to comparative information on hospitals’ quality of cancer care is limited. Patients’ interest in using this information when selecting a hospital for cancer surgery and the specific data they would desire are unknown. This study gauges patients’ demand for comparative information on hospitals’ quality of cancer surgery.

Methods:

We conducted a cross-sectional, national survey of 3,334 US residents who had received cancer surgery. The outcomes were patients’ reported likelihood of using a list of best hospitals for cancer surgery and patients’ reported interest in information about specific clinical outcomes, including 4-year survival after surgery, 30-day mortality after surgery, and rate of complications from surgery.

Results:

Two thirds of patients (68%) reported being actively involved in selecting a hospital for their surgery, and two thirds (65%) reported that their physician was involved in or made this decision. When asked what information might have helped them to choose a hospital, participants identified the hospital’s reputation (55%), patient satisfaction (44%), and the number of cancer surgeries performed at the hospital (36%). Approximately three quarters (73%) reported being likely to use a list of best hospitals for cancer surgery when selecting a hospital. Approximately 40% expressed interest in having information on at least one clinical outcome.

Conclusion:

Widespread interest exists among patients with cancer for comparative information on hospital quality as well as on clinical outcomes and hospitals’ reputation for cancer surgery. Policy reforms and additional research should address the unmet need for transparent, comprehensive data on the quality of hospitals’ cancer care.

INTRODUCTION

The Institute of Medicine identified information transparency, such as public reporting of quality and performance data, as a critical tool for improving the quality of health care.¹ Across a range of medical settings, providers’ quality of care and patient outcomes have been reported, including mortality rates after cardiac surgery^2,3; complications after joint replacements, prostate resections, spinal procedures, and gallbladder removals⁴; and readmissions after hospitalization for pulmonary diseases.⁵ Several studies also have demonstrated that public reporting has a positive effect on quality measures and outcomes, although findings have been mixed.^2,6-9 By providing transparent performance data, providers become accountable to the public, and patients, in theory, can use the information to make informed decisions about the most appropriate places to seek care.

Despite its importance, the amount of information available to patients who seek treatment for cancer is sparse.¹⁰ Some programs, such as the ASCO Quality Oncology Practice Initiative and the American College of Surgeons Commission on Cancer, provide data and comparison reports to individual providers for internal quality improvement purposes, but the data are not publicly reported and are inaccessible to patients.^11,12 Efforts to report the performance of hospitals for cancer care at the state and federal levels are limited in scope. Only a few states have released data on their hospitals’ volume and select measures of cancer care quality.^13-15 At the federal level, the Affordable Care Act mandates prospective payment system–exempt cancer hospitals to report publicly select measures of quality,¹⁶ but this applies to only 11 hospitals and includes < 20 quality measures. Although these hospitals deliver specialized cancer care, most cancer care is delivered in community settings.^17,18 Thus, none of these efforts establish a comprehensive resource for patients with complete information for all US hospitals that deliver cancer care across the nation.

Moreover, measures selected for reporting do not necessarily reflect information that patients wish to have, which is a barrier to the use of public reporting efforts. Patients generally express interest in having access to information on care quality^19-23 and respond positively to efforts to release comparative information.²⁴ However, little research has focused specifically on patients with cancer and the information they want to help to decide among hospitals for their care.^20,25-28

Whether patients with cancer would use information on the quality of hospital care to make a decision about where to go for their care and the type of information they are interested in are not well understood. To shed light on these crucial issues, we embarked on a cross-sectional national survey of patients with cancer. Our objective was twofold: To evaluate patients’ interest in using a list of best hospitals for cancer surgery and to determine the type of information patients would want to have about hospital quality. We hypothesized that patients would be interested in published comparative information on hospital performance and would highly value information on clinical outcomes.

METHODS

Study Design

This study analyzes data from a cross-sectional Web-based survey developed in collaboration with ProPublica (New York, NY), an investigative newsroom, and Inspire, an online patient support community. We followed recommended practices for Internet surveys according to the Checklist for Reporting Results of Internet E-Surveys statement²⁹ (Data Supplement). This checklist is analogous to other recommendations, such as the CONSORT statement for randomized trials.

Sample

Participants were drawn from the Inspire patient community, an online network of patients and caregivers that represents > 3,000 conditions across > 200 disease-specific groups. The survey was designed to take approximately 10 minutes. Inspire sent the survey to all individuals who had previously opted to receive health-focused market research opportunities and were members of select communities that represent the following cancer types: breast, bladder, colorectal, cervical, kidney, ovarian, prostate, stomach, thyroid, and testicular. The invitation also was distributed to two general cancer communities. No incentive was provided for participation. The survey invitation was e-mailed to approximately 182,000 accounts; 38,995 e-mail invitations were opened, 5,423 people clicked on the online survey link provided in the e-mail invitation, and 4,714 people completed the survey (3,334 surveys completed by patients and 1,380 by caregivers). We report the responses from patients only.

Survey Administration

An e-mail request was sent to eligible Inspire accounts on January 19, 2017, that asked for participation in the survey. An e-mail reminder was sent to all nonrespondents on February 6, 2017. Data collection was completed on February 10, 2017.

Survey Instrument

We adapted questions from prior studies^25,30,31 and added new questions that focused on the study objectives (Data Supplement). We asked patients to report demographic characteristics (age, sex, race, state, education level) and about how they selected a hospital for surgery (eg, whether they considered other hospitals and tried to find comparative information, who was involved in making the decision) and their satisfaction with hospital care.

We included two main questions of interest: “ProPublica is considering publishing a list of the best hospitals based on the quality of their cancer surgery. If you had access to this list when making the decision about which hospital to choose for cancer surgery, how likely would you be to use this list?” and “What information about hospitals would you have wanted before you made the decision about which hospital to choose for cancer surgery?” Patients could select up to four of the following options: whether the hospital is in their insurance network, out-of-pocket cost for the surgery, hospital reputation for cancer surgery, hospital reputation for any type of care, patients’ satisfaction with care, whether all patients are treated equally, how quickly patients receive surgery, volume of surgeries, how common it is for patients to have complications from surgery, how common it is for patients to die as a result of surgery within 30 days of surgery, and how common it is for patients to be alive 4 years after surgery. Response options were randomized across patients.

Outcomes

The outcomes of interest were patients’ reported interest in a list of best hospitals for cancer surgery and patients’ reported interest in information on hospitals’ clinical outcomes. To determine interest in clinical outcomes, we created a binary variable for wanting information on at least one of three clinical outcomes: how common it is for patients to have complications from surgery; how common it is for patients to die as a result of surgery within 30 days of surgery; and how common it is for patients to be alive 4 years after surgery.

Patient Characteristics

We were interested in potential differences in outcomes by patient characteristics, including demographic information (sex, age, region, race, education), cancer type, patient or physician involvement in the decision to select a hospital, and satisfaction with the surgery experience. We categorized cancer types as common versus rare on the basis of the National Cancer Institute list of common cancers.²² For those who selected multiple cancers, respondents were categorized as common if at least one of the selected cancer types was common.

Analysis

We used descriptive statistics to address the study objectives and multivariable logistic regression to test the associations between patient characteristics and the outcomes of interest. All analyses were performed using R version 3.3.3 statistical software (The R Project for Statistical Computing, https://cran.r-project.org/bin/windows/base/old/3.3.3). The study was exempt from review by the institutional review board of Memorial Sloan Kettering Cancer Center because the data were anonymized.

RESULTS

Patient Characteristics

Our sample included 3,334 US patients who had received surgery for cancer. Patient characteristics are listed in Table 1. Patients who completed the survey had similar characteristics to the larger Inspire community. The majority (69%) were ≥ 50 years of age. Respondents were predominantly female (71%) and white (91%), and most had a bachelor’s degree or higher (66%). Three quarters (75%) were diagnosed with cancer within the past 5 years. The most common cancer diagnoses were thyroid (20%), breast (18%), bladder (16%), ovarian (15%), and prostate (10%). Approximately 15% reported having two or more cancers. Patients were geographically diverse and represented all 50 states.

Table 1.

Characteristics of Patient-Respondents Who Had Received Surgery for Cancer

Decision Making Around Selecting a Hospital for Surgery

Approximately one half (49%) of the patients reported that there was a closer hospital that offered cancer surgery than the one they selected, which suggests that a large number of patients do not select a hospital purely on convenience (Table 2). Twenty-two percent of respondents seriously considered receiving their surgery elsewhere. The majority of patients were involved in selecting a hospital for surgery (68%) and were satisfied with the quality of their surgical experience (86%). Approximately 65% reported that their physicians were involved in these decisions. Physicians were the primary decision makers for approximately one half of these patients.

Table 2.

Responses to Questions About Selecting a Hospital for Cancer Surgery

Likelihood of Using a List of Best Hospitals

The majority of respondents (73%) reported that they would be somewhat or very likely to use a list of best hospitals (Fig 1). Slight differences by patient characteristics were observed (Data Supplement). Older patients were less likely to report interest in using a list (66% age > 70 years v 79% age ≤ 50 years), whereas those who had been involved in selecting the hospital were slightly more likely to report interest in using a list (76%) than those for whom physicians were the primary decision makers (67%). In adjusted analyses, being involved in selecting the hospital was associated with a higher odds of being likely to use a list (odds ratio [OR], 1.53; 95% CI, 1.29 to 1.81).

Fig 1.

Likelihood of using a list of best hospitals for cancer surgery and information patients would have wanted to know about hospitals. The two main questions of interest were: “ProPublica is considering publishing a list of the best hospitals on the basis of the quality of their cancer surgery. If you had access to this list when making the decision about which hospital to choose for cancer surgery, how likely would you be to use this list?” and “What information about hospitals would you have wanted before you made the decision about which hospital to choose for cancer surgery? Please select up to four options.” Out-of-pocket cost for the surgery and whether the hospital was in the patient’s insurance network were excluded from the analyses to focus on quality-related factors.

Information Desired Before Selecting a Hospital for Surgery

One quarter (27%) of patients tried to find information that compared the selected hospital with another hospital before making a decision (Table 2). When asked what they would have wanted to know when selecting a hospital for cancer surgery, patients were most interested in the hospital’s reputation for cancer surgery (55%) followed by patient satisfaction (44%), volume of surgeries performed at the hospital (36%), and hospital’s reputation for any type of care (33%; Fig 1). Minority patients (48%) were less likely than white patients (55%) to desire information on the hospital’s reputation for cancer surgery, but minority patients were more likely (12%) than white patients (6%) to desire information on whether hospitals treat all patients equally.

We determined the unadjusted rates for patients’ reported interest in data on at least one of three clinical outcomes: 4-year survival, 30-day mortality after surgery, and complications from surgery (Data Supplement). Approximately 40% expressed interest in at least one of these clinical outcomes when selecting a hospital for cancer surgery. Overall, slight differences were found in respondents’ interest in clinical outcomes by their characteristics. Older respondents, nonwhite respondents, and respondents with less than a bachelor’s degree were somewhat less likely to report interest in clinical outcomes (32% > 70 years of age v 44% ≤ 50 years of age; 32% nonwhite v 41% white; 35% with less than bachelor’s degree v 42% with a bachelor’s degree or higher). Respondents who were dissatisfied with their past care were slightly more interested in clinical outcomes (46%) than those who were satisfied (39%). In adjusted analyses, nonwhite respondents (OR, 0.593; 95% CI, 0.43 to 0.81) and those satisfied with their care (OR, 0.72; 95% CI, 0.58 to 0.89) showed slightly lower odds of wanting to know clinical outcomes (Data Supplement). Having a bachelor’s degree or higher was associated with slightly higher odds of wanting to know at least one clinical outcome (OR, 1.34; 95% CI, 1.14 to 1.23).

DISCUSSION

The findings of this national survey demonstrate a strong and widespread interest among patients with cancer in having access to comparative information on the quality of hospitals’ cancer surgery. In addition, many patients in this study reported being interested in measures of specific clinical outcomes and hospitals’ reputation for cancer surgery.

We specifically asked patients across multiple cancer types and ages about their interest in a list of hospitals for cancer surgery. We observed high levels of interest, with 73% of respondents reporting that they would be likely to use a list of best hospitals. This finding is largely consistent across patient characteristics and shows that patients of all ages and education levels and with a wide variety of cancer types across the entire United States desire cancer-specific comparative information. This finding is also consistent with other studies that addressed similar issues for other clinical domains. For example, on the basis of a survey provided to visitors of existing online public reporting Web sites, consumers reported that they want more information about how well hospitals treat specific conditions, including cancer.²¹ When Schwartz et al²⁵ asked 510 Medicare beneficiaries whether they would use a list of hospitals for various surgical operations, many stated that they might refer to such a list. In our survey of > 3,000 patients with cancer active in an online support community, three quarters reported that they would be likely to use a list of hospitals for cancer surgery. These findings emphasize the unmet need for cancer-specific comparative information for patients with cancer.

Patients who responded to our survey reported that they would have used information on clinical outcomes and other quality measures when selecting a hospital for surgical cancer treatment. Clinical outcomes (eg, long-term survival, 30-day mortality after surgery) are the standards by which providers are evaluated, and we hypothesized that patients also would value these measures of clinical quality. In our study, many patients (40%) expressed interest in specific clinical outcomes, which aligns with previous studies that have shown that patients are interested in global measures of clinical outcomes.^20,26 We also observed that patients highly value having information on the hospital’s reputation for cancer surgery (55%). However, patients’ conceptualization of reputation is not well understood and possibly encompasses quality-related factors, such as volume and clinical outcomes. Understanding this conceptualization will help to ensure that publicly reported data are meaningful to patients.

The current findings build on existing studies that have shown differences in decision making by race. For example, Freedman et al³¹ surveyed 500 patients with breast cancer in California about selecting hospitals and surgeons and showed that minority patients were less likely than white patients to report provider reputation as an important factor in their decision making. We similarly observed differences by race and found that minority patients were less likely to desire information on hospitals’ clinical outcomes and reputation for cancer treatment. Instead, we observed that minority patients were more likely than white patients to desire information on whether hospitals treat patients equally. These results suggest that sociodemographic factors should be taken into consideration when understanding the types of information important to patients.

Effective reporting of performance data may require making the information meaningful to both patients and providers. Patients in our study reported that referring physicians played a significant role in helping them to select a hospital for cancer surgery. This finding is consistent with other research; for example, Lux et al³² found that physicians in private practice were important sources of information for gynecologic oncology patients. More research is needed to ascertain who the referring physicians are for cancer surgery (eg, primary care physicians, medical oncologists); how they decide which hospitals to send their patients to; and what performance data, if any, they take into consideration when making these decisions. If physicians are a key audience for publicly reported data, including their input during the development of these efforts would likely increase their use.

Although this study provides evidence of an unmet need for comparative information on hospital quality, some limitations exist. Data were self-reported. Although the survey respondents may have indicated that they are interested in using information, whether or how they would use publicly reported information in a real-world setting is unclear. The respondents represented a convenience sample. Approximately 39,000 Inspire members (21%) opened the e-mail invitation, and > 4,000 patients and caregivers (3%) completed the survey. Furthermore, survey respondents may not be representative of the overall population of patients who seek surgical treatment of cancer. Male patients and minority patients were less represented among the respondents compared with the distribution of all surgical patients with cancer. Thus, caution should be used in generalizing respondents’ opinions to patients beyond active users of online support communities who are seeking surgical treatment. Additional study is needed to gather insights from other patients who were not as well represented in this study, including minorities and nonusers of online patient support communities, about their perceptions of the usefulness of this type of information in their decision making. Nonetheless, the study captured a wide range of users and experiences from a large sample of > 3,000 patients with diverse cancers and geographical locations across all 50 US states. These respondents likely represent the group most likely to seek out publicly reported comparative information, so having obtained their input is valuable.

At the policy level, opportunities may exist to promote transparency in cancer care quality by expanding reporting requirements. One approach would be to broaden current reporting requirements that apply only to the 11 prospective payment system–exempt hospitals to all hospitals across the nation that deliver cancer care and to include other measures of quality that are important and are of interest to patients. If reporting was voluntary, it could be incentivized. Such changes would help to address the need for transparent and comparative information on hospitals’ quality of cancer care and lead to more-informed patient choice.

In conclusion, an unmet need exists for comprehensive information on hospitals’ quality of cancer care. Patients with cancer express a strong interest in having comparative information for selecting hospitals and have demonstrated interest in having data on clinical outcomes and hospitals’ reputation for cancer surgery. More work is needed to ensure that the void for performance data is filled with information on cancer care quality that is meaningful and useful for patients.

AUTHOR CONTRIBUTIONS

Conception and design: Peter B. Bach, David J. Taylor, Allison Lipitz-Snyderman

Administrative support: Annie Yang

Collection and assembly of data: Annie Yang, David J. Taylor, Allison Lipitz-Snyderman

Data analysis and interpretation: Annie Yang, Susan Chimonas, David J. Taylor, Allison Lipitz-Snyderman

Manuscript writing: All authors

Final approval: All authors

Accountable for all aspects of work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Critical Choices: What Information Do Patients Want When Selecting a Hospital for Cancer Surgery?

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jop/site/ifc/journal-policies.html.

Annie Yang

No relationship to disclose

Susan Chimonas

No relationship to disclose

Peter B. Bach

Leadership: ExamWorks

Honoraria: American Society for Health-System Pharmacists, Gilead Sciences, WebMD, Goldman Sachs, Defined Health, Vizient, Anthem, Excellus BlueCross BlueShield, Hematology/Oncology Pharmacy Association, Novartis, Janssen Pharmaceuticals, Third Rock Ventures, JMP Securities

Consulting or Advisory Role: Foundation Medicine

Research Funding: Kaiser Permanente, Laura and John Arnold Foundation

Travel, Accommodations, Expenses: Bipartisan Congressional Health Policy Conference, Yale, American Society for Health-System Pharmacists, Gilead Sciences, WebMD, Goldman Sachs, Defined Health, Vizient, Foundation Medicine, Anthem, Excellus BlueCross BlueShield, Hematology/Oncology Pharmacy Association, Novartis, Janssen Pharmaceuticals, Third Rock Ventures

David J. Taylor

No relationship to disclose

Allison Lipitz-Snyderman

No relationship to disclose