Potential Bias Found in Patient Payer Type Affecting Satisfaction Surveys

Patient satisfaction is becoming an increasingly important factor in healthcare, affecting both clinical management and reimbursement of physicians and other providers. Satisfaction surveys originated in the 1980s for internal process improvement(1); since then, they have evolved into assessment tools for hospital and physician performance using the Hospital Consumer Assessment of Healthcare Providers and Services (HCAHPS) survey and the Clinician and Group Consumer Assessment of Healthcare Providers and Services (CG-CAHPS) survey, respectively. The CAHPS surveys currently are used by CMS as part of the reimbursement algorithm.(2)

In 2015, CMS established the Medicare Access and CHIP Reauthorization Act (MACRA), which foundationally changed the way physicians are reimbursed.(3) Instead of the traditional fee-for-service model, MACRA instituted a four-part assessment of physician performance called the Merit-Based Incentive Payments System (MIPS), which includes the following metrics: (1) quality; (2) Advancing Care Information; (3) clinical practice improvement activities; and (4) resource use.(4) However, the four metrics of the MIPS system are not weighted equally. The quality component accounts for 60% of a physician’s final score.(5) The largest component of the quality metric is patient satisfaction, giving the various CAHPS surveys significant power in reimbursement adjustments.(5)

A key construct in the MIPS system is that it is budget-neutral, penalizing physicians who perform worse than the national median and rewarding physicians who perform better than the national median.(6) The maximum penalty a low-performing physician can incur is a 9% deduction in CMS reimbursement. Similarly, a top-performing physician can receive up to a 9% bonus in CMS reimbursement.(2)

Because a significant portion of reimbursement is dependent on patient satisfaction, investigation is warranted to determine whether any variables outside of physician or hospital performance influence the way patients rate their experiences. A global literature review done in March 2017 isolated several determinants of patient satisfaction, both healthcare provider–related elements and also patient-related characteristics, such as age, gender, education, socioeconomic status, and geographical characteristics.(7) Overall, the researchers determined that “there is a need for more studies on how behavioral, cultural, and socio-demographic differences affect patient satisfaction.”(7)

Given the limited exploration into the drivers of patient satisfaction and how they specifically affect survey scores, our present aims to evaluate the role of insurance type on patient satisfaction in an academic OB/GYN setting.

Materials and Methods

We performed a cross-sectional, retrospective study investigating self-reported, deidentified patient satisfaction surveys completed from all OB/GYN clinics affiliated with a large university in the United States. All patients who were seen by a billable provider were screened for inclusion in the study. Inclusion criteria consisted of completion of a Press Ganey survey, either electronic or paper, between December 1, 2016 and November 30, 2017. The Press Ganey system at our institution sends an e-mail survey to all arrived patients at a 100% send rate. A paper survey is sent randomly to 1 out of 5 arrived appointments. Press Ganey patient satisfaction surveys are scored on a scale of 0 to 100 points. Data were extracted from the Press Ganey surveys to include mean patient satisfaction survey scores plus or minus standard deviation and total number of surveys for each coded insurance plan.

Primary insurance information, extracted from patients’ charts rather than self-reported, also was collected via Press Ganey. The various insurance plans were grouped into five categories: commercial; government; Medicaid; Medicare; and self-pay. Government plans included any insurance issued by a government entity that is not considered Medicare or Medicaid, such as Veterans Administration benefits provided by Veterans Affairs and coverage provided by the State Department of Criminal Justice. Patients who carry primary and secondary insurance were grouped by their primary insurance; further categorization by secondary insurance was not available from the Press Ganey system. The business offices at our institution calculated the percentage of Medicare patients secondarily covered by Medicaid who were seen at affiliated OB/GYN clinics during the timeframe of this study.

After weighted mean satisfaction scores plus or minus standard deviation were calculated for the five groups, a one-way analysis of variance (ANOVA) was conducted to determine whether the patient satisfaction scores were significantly different between the groups. Post-hoc analysis was performed using the Tukey-Kramer Honestly Significant Difference method to compare individual groups against each other and control for familywise error rate. A 95% confidence interval was included within the results of the Tukey-Kramer test. All statistical testing was done with an alpha level of 0.05.

Results

For the study, 5033 patients met inclusion criteria, and their deidentified patient satisfaction surveys underwent analysis. The 5033 surveys analyzed fell into these insurance groups:

• Commercial insurance: 2505 patients (50%);

• Medicare: 1801 patients (36%);

• Medicaid: 454 patients (9%);

• Other government insurers: 261 patients (5%); and

• Self-pay: 12 patients (0.24%).

Seventy-four insurance plans were represented in the data. Commercial insurance had the greatest variety of plans (32), followed by Medicaid (17), government plans (11), Medicare (7), and self-pay (4).

Out of 100 possible points on the Press Ganey survey report, patients with Medicare reported the highest mean satisfaction score, 93.31 ± 1.44. This was followed by patients with other government insurance (92.8 ± 2.41), commercial insurance (91.91 ± 1.26), and self-pay (90.88 ± 4.36). The lowest mean score was reported from patients with Medicaid (88.85 ± 2.01; Table 2).

ANOVA analysis revealed an F value of 852.8 (p <.001). Post-hoc analysis showed the following results:

• Medicaid versus Medicare produced the largest difference between groups, with a difference of 4.46 (CI: 4.24-4.67; p <.001).

• Government versus Medicare produced the smallest difference, of 0.51 (CI: 0.23-0.78; p <.001).

• All other between-group differences are as follows:

  • Commercial versus government, 0.89 (0.62-1.16; p <.001);

  • Commercial versus Medicaid, 3.06 (CI: 2.85-3.27; p <.001);

  • Commercial versus Medicare, 1.40 (CI: 1.27-1.53; p <.001);

  • Commercial versus self-pay, 1.03 (CI: –0.15-2.21; p = .121);

  • Government versus Medicaid, 3.95 (CI: 3.63-4.26; p <.001);

  • Government versus self-pay, 1.92 (CI: 0.71-3.13; p <.001);

  • Medicaid versus self-pay, 2.03 (CI: 0.84-3.22; p <.001); and

  • Medicare versus self-pay, 2.43 (CI: 1.24-3.61; p <.001).

With the exception of commercial vs self-pay, all intergroup comparisons show a statistically significant difference. These Tukey-Kramer results are outlined in Table 3.

Discussion

Our results show that there is a statistically significant difference between almost all payer types as it relates to patient satisfaction survey scores. This is in line with a 2014 study, which found that insurance type is one of the top three factors contributing to a patient’s satisfaction with healthcare.(8) The exception to our results is commercial versus self-pay, whose scores were not found to be statistically significantly different. This could be explained by a similar patient demographic in those two groups.

Patients covered by Medicare often have secondary insurance. The Press Ganey system at our institution is unable to delineate secondary insurance coverage for patients. However, the business offices of our institution, which have the capability to examine secondary insurance coverage, report that during the time frame of this study about 35% of all patients with Medicare coverage seen at affiliated clinics were secondarily covered by Medicaid. Because the Press Ganey system was unable to account for secondary insurance coverage, any individual with dual coverage was grouped within their primary insurance class. The overlap of insurance by Medicare patients who also carry Medicaid may falsely reduce the extent of the true difference between the two groups. We suggest that the actual difference is most likely larger than we are able to report and warrants further investigation in future studies.

Government insurance had the next highest satisfaction scores. Most of the survey responses in this group were patients with VA and military coverage, suggesting higher patient satisfaction among active military and veterans.

In Texas, where Medicaid has not expanded, a pregnant woman qualifies for Medicaid coverage if her household income is lower than certain standards (e.g., $2106 per month if she is alone or $5063 month for a family of five).(9) However, this coverage ends after two months postpartum. Since eligibility is based largely on income and tends to be a shorter-term solution for insurance coverage, particularly for pregnant women, this could result in a high turnover of Medicaid patients and lower satisfaction. After the postpartum period, the woman may enroll in a separate Medicaid program called Healthy Texas Women (HTW) Medicaid, which serves eligible women of reproductive age.

The systematic challenge for these women is that although perinatal Medicaid provides comprehensive coverage, HTW has limited benefits, covering only the following:

• Pregnancy testing;

• Pelvic examinations;

• Sexually transmitted infection services;

• Breast and cervical cancer screenings;

• Clinical breast examination;

• Mammograms;

• Screening and treatment for cholesterol, diabetes and high blood pressure;

• HIV screening;

• Long-acting reversible contraceptives;

• Oral contraceptive pills;

• Permanent sterilization;

• Other contraceptive methods such as condoms, diaphragm, vaginal spermicide, and injections; and

• Screening and treatment for postpartum depression.(10)

The limited coverage provided by HTW poses two main concerns regarding patient satisfaction in an OB/GYN setting. The first relates to women who alternate between perinatal Medicaid and HTW Medicaid coverage throughout their reproductive years. The stark difference in coverage between these entities can lead to confusion and overall dissatisfaction for these patients. The second concern is more general and relates to all eligible women of reproductive age, who would understandably be less satisfied if their medical needs were not covered by the limited benefits of HTW Medicaid. It is likely that patients in other states with varying Medicaid coverage experience similar dissatisfaction.

Overall, our results agree with studies of similar scope. Liu et al.(11) found hospitals with more Medicaid patients received lower patient satisfaction scores than hospitals with fewer Medicaid patients. Bible et al.,(12) in their study in an outpatient spine clinic, observed that patients covered by Medicare were the most satisfied with their provider.

We acknowledge limitations to this study. First, those without email access, such as more senior patients, were excluded. Although Press Ganey mails a random paper survey to one in five patients, email is the predominant source of satisfaction survey collection and, therefore, could lead to inherent selection bias. A second study limitation relates to the self-pay category, which had a small number of survey responses (n=12). This introduces more variance into comparisons between self-pay and the other four categories and limits the conclusions that can be drawn about patient bias. Third, it has been well studied that those responding to electronic surveys differ from the population invited to participate in the survey, suggesting nonresponse bias and selection bias.(13) Another study found that respondents to patient satisfaction surveys were more likely to report a more positive hospital experience, whereas patients experiencing negative thoughts about their hospital stay were less likely to complete a patient satisfaction survey.(14) Selection bias and non-response bias potentially may skew our data.

Implications for Practice and Policy

We recommend that current practitioners be mindful that insurance type plays a role in overall patient satisfaction. Physicians and other healthcare professionals should aim to increase their understanding of the challenges that patients face, especially those on limited insurance plans. Knowing a patient has limited coverage may affect the way treatment options are presented so that the patient understands what she is going to be financially responsible for.

Our data indicate that a policy change in the algorithm for MIPS reimbursement may be useful to account for the factors that affect patient satisfaction. Further research is needed to identify the nonmodifiable variables and their respective effects on satisfaction scores. These studies would aid in determining any adjustments needed to ensure equitable reimbursement policy.

Conclusions

Our data show a trend toward patient satisfaction scores being linked to insurance coverage type. This is in line with current literature indicating that patient satisfaction is multifaceted in nature. Our data warrant further studies of individual, patient-level data that may lead to the adjustment of the way patient satisfaction is evaluated and reimbursed, recognizing that survey scores are affected by more than provider performance alone.

Acknowledgment: This study was supported in part by the TTUHSC Clinical Research Institute.

References

1. Kash B, McKahan M. The evolution of measuring patient satisfaction. J Prim Heal Care Gen Pract. 2017;1(1):1-4. www.scientonline.org/open-access/the-evolution-of-measuring-patient-satisfaction.pdf . Accessed July 20, 2020.

2. Medicare Program; Merit-Based Incentive Payment System (MIPS) and Alternative Payment Model (APM) Incentive Under the Physician Fee Schedule, and Criteria for Physician-Focused Payment Models. Final rule with comment period. Fed Regist. 2016;81(214):77008-77831. www.federalregister.gov/documents/2016/11/04/2016-25240/medicare-program-merit-based-incentive-payment-system-mips-and-alternative-payment-model-apm . Accessed May 12, 2020.

3. Mullins A. Medicare payment reform: making sense of MACRA. Fam Pract Manag. 2016;2016(March-April):12-15. www.ncbi.nlm.nih.gov/pubmed/26977983 . Accessed May 12, 2020.

4. Miller P, Mosley K. Physician reimbursement: from fee-for-service to MACRA, MIPS and APMs. J Med Pract Manage. 2016;31:266-269. Accessed May 12, 2020.

5. Hirsch JA, Rosenkrantz AB, Ansari SA, Manchikanti L, Nicola GN. MACRA 2.0: Are you ready for MIPS? J Neurointerv Surg. 2017;9(7):713-715. doi:10.1136/neurintsurg-2016-012845

6. Adler KG. Holy MACRA! Will our future be better or worse? Fam Pract Manag. 2016;23(2):5.

7. Batbaatar E, Dorjdagva J, Luvsannyam A, Savino MM, Amenta P. Determinants of patient satisfaction: a systematic review. Perspect Public Health. 2017;137(2):89-101. doi:10.1177/1757913916634136

8. Deshpande SP, Deshpande SS. Factors influencing consumer satisfaction with health care. Health Care Manag (Frederick). 2014;33(3):261-266. doi:10.1097/HCM.0000000000000024

9. Health Care—How to Get Help. https://yourtexasbenefits.hhsc.texas.gov/programs/health/women/pregnant. Accessed July 1, 2020.

10. HTW: Benefits. Healthy Texas Women. www.healthytexaswomen.org/healthcare-programs/healthy-texas-women/htw-benefits . Accessed July 20, 2020.

11. Liu SS, Wen YP, Mohan S, Bae J, Becker ER. Addressing Medicaid expansion from the perspective of patient experience in hospitals. Patient. 2016;9:445-455. doi:10.1007/s40271-016-0167-y

12. Bible JE, Kay HF, Shau DN, O’Neill KR, Segebarth PB, Devin CJ. What patient characteristics could potentially affect patient satisfaction scores during spine clinic? Spine (Phila Pa 1976). 2015;40:1039-1044. doi:10.1097/BRS.0000000000000912

13. Compton J, Glass N, Fowler T. Evidence of selection bias and non-response bias in patient satisfaction surveys. Iowa Orthop J. 2019;39(1):195-201.

14. Perneger TV, Chamot E, Bovier PA. Nonresponse bias in a survey of patient perceptions of hospital care. Med Care. 2005;43:374-380. doi:10.1097/01.mlr.0000156856.36901.40