Orthopedics

Feature Article Supplemental Data

The Effect of Length of Hospital Stay and Patient Factors on Patient Satisfaction in an Academic Hospital

Waqaar Diwan, MD; Paul A. Nakonezny, PhD; Joel Wells, MD, MPH

Abstract

The Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) is a metric for patient satisfaction consisting of 19 questions divided into 10 domains. Scores affect hospital reimbursements and accreditation and may play a role in patient outcomes. It is unclear how length of stay and other factors affect each of the 10 domains. This retrospective review gathered data of 600 patients between December 2008 and January 2017 who completed the HCAHPS survey. The odds of complete satisfaction in each of the 10 domains was evaluated. The results suggest increased length of stay is associated with lower odds of patient satisfaction and decreased likelihood of recommending the hospital. The odds of being completely satisfied regarding communication with physicians, discharge information, and responsiveness of the hospital staff, as well as the odds of recommending the hospital to others, were lower if the care provider was younger than the patient. Obese patients were also more likely to be satisfied with responsiveness and care transition. Male patients were more satisfied with communication about medications (odds ratio [OR], 1.694), care transition (OR, 1.489), and cleanliness (OR, 2.120). Medicare and fewer consults were related to increased odds of patient satisfaction with care transition (OR, 1.748 and 0.573, respectively). Males, older patients, and White patients were more likely to recommend the hospital (OR, 1.476, 1.025, and 1.690, respectively). Length of stay affects patient satisfaction and likelihood of recommending the hospital to others. Other factors such as a younger provider age than the patient, lower body mass index, female sex, non-Medicare insurance, and higher number of consults are also associated with lower satisfaction in various domains. Hospital systems can bolster patient satisfaction by strategizing day-of-surgery and weekend staffing to reduce length of stay. [Orthopedics. 2020;43(6):373–379.]

Abstract

The Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) is a metric for patient satisfaction consisting of 19 questions divided into 10 domains. Scores affect hospital reimbursements and accreditation and may play a role in patient outcomes. It is unclear how length of stay and other factors affect each of the 10 domains. This retrospective review gathered data of 600 patients between December 2008 and January 2017 who completed the HCAHPS survey. The odds of complete satisfaction in each of the 10 domains was evaluated. The results suggest increased length of stay is associated with lower odds of patient satisfaction and decreased likelihood of recommending the hospital. The odds of being completely satisfied regarding communication with physicians, discharge information, and responsiveness of the hospital staff, as well as the odds of recommending the hospital to others, were lower if the care provider was younger than the patient. Obese patients were also more likely to be satisfied with responsiveness and care transition. Male patients were more satisfied with communication about medications (odds ratio [OR], 1.694), care transition (OR, 1.489), and cleanliness (OR, 2.120). Medicare and fewer consults were related to increased odds of patient satisfaction with care transition (OR, 1.748 and 0.573, respectively). Males, older patients, and White patients were more likely to recommend the hospital (OR, 1.476, 1.025, and 1.690, respectively). Length of stay affects patient satisfaction and likelihood of recommending the hospital to others. Other factors such as a younger provider age than the patient, lower body mass index, female sex, non-Medicare insurance, and higher number of consults are also associated with lower satisfaction in various domains. Hospital systems can bolster patient satisfaction by strategizing day-of-surgery and weekend staffing to reduce length of stay. [Orthopedics. 2020;43(6):373–379.]

The Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey was developed to allow comparison of hospitals and provide a standardized way to rate hospitals, promote hospital improvements, and bring accountability and transparency to hospital care.1 In 2012, the Centers for Medicare & Medicaid Services (CMS) began linking HCAHPS scores to hospital reimbursement.2 The CMS calculates a composite quality score, with the HCAHPS representing approximately 40% of this score.2,3 The HCAHPS and its effect on hospital bottom line generate criticism, but studies show the survey has a direct relationship with patient outcomes and quality of care.4–6 Tsai et al5 found that high patient satisfaction correlated with important hospital metrics such as process quality, lower readmission and mortality rates, and shorter length of stay (LOS). There exists a paucity of research evaluating associations of various patient and provider factors with patient satisfaction of care using the HCAHPS domains.

One factor that may play a role in patient satisfaction is LOS. Reducing LOS has been studied extensively regarding the minimization of nosocomial infections and hospital costs.7,8 However, the effect of LOS on various HCAHPS domains of satisfaction has not been previously studied.7 The purpose of this study was to evaluate the relationship between LOS and patient satisfaction in each of the 10 domains of the HCAHPS survey. The authors also evaluated the relationship between patient and provider factors and the various domains of patient satisfaction. The results of this study will inform providers regarding the factors that affect satisfaction and, ultimately, bolster patient outcomes.

Materials and Methods

Participants

The authors reviewed satisfaction scores from the hospital encounters of 646 patients at a single academic medical center between December 2008 and June 2017. All clinical encounters with patients 18 years and older who completed the HCAHPS survey following a hospital stay were included in the study. After obtaining patient responses to the HCAHPS survey, the following information was extracted from the patients' medical records for the purpose of this cross-sectional study: age, sex, race, religion, marital status, comorbidities, type of health insurance, body mass index (BMI), provider, and type of orthopedic hospital visit. A Charlson Comorbidity Index (CCI) was calculated for each patient.9 Information on provider age and race was also obtained and included for analysis. This study received institutional review board approval.

Outcome Variable

The HCAHPS survey consists of 19 questions divided into 10 domains: communication with nurses, communication with physicians, responsiveness of hospital staff, communication about medicines, discharge information, care transition, cleanliness of hospital environment, quietness of hospital environment, global hospital rating, and recommend the hospital. To focus on various aspects of patient satisfaction with the hospital encounter, the primary outcomes of patient satisfaction were the 10 domains of the HCAHPS survey. The items for each domain were measured on a 4-point Likert-type scale and the response categories were: never, sometimes, usually, and always. Some of the domains included a qualifying yes or no question to evaluate whether the measure applied to the patient's experience. The hospital rating measure had 10 options, 0 (worst hospital possible) to 10 (best hospital possible). The answer choices were scored using a top-box methodology as defined by the CMS10 and were summed to produce a total score for each domain that ranged from 0 to 100. A higher total score indicated greater satisfaction within the specific HCAHPS domain. A detailed explication of the scoring methodology is outlined by the CMS.10

Because the authors observed a high frequency distribution (range, 51%–87%) on the 10 HCAHPS domains for which patients were completely satisfied (HCAHPS score, 100), patient satisfaction was operationalized as a binary outcome: completely satisfied vs not completely satisfied. Patients were categorized as completely satisfied if their specific HCAHPS domain (subscale) score was 100. Otherwise, patients with a specific HCAHPS domain (subscale) score less than 100 were categorized as not completely satisfied. This threshold was chosen a priori and is consistent with the operational definition used by Martin et al.11 The authors modeled the probability of the patient being completely satisfied.

Potential Predictor Variables

An initial pool of 14 characteristic variables was selected for analysis as potential predictors of patient satisfaction. These variables were selected based on the results of previously published findings.11–19 The pool of potential predictors, which was selected a priori, included patient age (years), patient sex (male vs female), care provider is younger in age than patient (yes vs no), patient race (White vs non-White), care provider is same race as patient (yes vs no), patient marital status (married vs not married), number of previous patient consults (0, 1, 2 to 5; with 0 as the reference group), patient BMI (≥30 kg/m2 vs <30 kg/m2), patient health insurance status (Medicare vs private), hospital admission due to a previous complication (yes vs no), LOS in the hospital (days), patient CCI, number of patient allergies, and arthroplasty orthopedic subspecialty (yes vs no).

Multiple Imputation for Missing Values

Missing values for the HCAHPS items and predictor variables, which occurred in no more than approximately 10% of the sample, were imputed. Missing values (with an assumed arbitrary missing pattern) for the classification variables and for the continuous variables were imputed via 500 burn-in iterations (samples) using fully conditional specification along with the discriminant method (for the classification variables) and the predictive mean matching method (for continuous variables) of the process of multiple imputation (PROC MI) procedures in SAS, version 9.4, software (SAS Institute, Cary, North Carolina).20

Statistical Analysis

Demographic and clinical characteristics for the sample of 646 orthopedic patients were described using the sample mean and standard deviation for continuous variables and the frequency and percentage for categorical variables. Multiple logistic regression, with penalized maximum likelihood estimation along with Firth's bias correction, was implemented to estimate the odds of the patient being completely satisfied on each of the 10 HCAHPS domain outcomes from the set of regressors (predictors). A separate logistic regression model was implemented for each of the 10 HCAHPS domain outcome measures. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were reported. An estimated OR greater than 1 indicated greater odds of the patient being completely satisfied. Statistical analyses were performed using SAS, version 9.4, software. The level of significance was set at alpha=0.05 (two-tailed).

Results

Participant Characteristics

Of the 646 orthopedic patients, 44.89% were male, 79.10% were non-Hispanic White, and mean patient age was 67.14±11.96 years. Mean BMI was 29.54±6.14 kg/m2, with 57.12% of patients having a BMI less than 30 kg/m2. Fifty-four percent of the orthopedic patients were arthroplasty cases, with a mean LOS of 2.57±2.55 days. Forty-eight percent of the patients had Medicare insurance. Demographic and clinical characteristics of the overall sample of orthopedic patients, including mean HCAHPS patient satisfaction scores, are shown in Table 1.

Demographic and Clinical Characteristics of the Overall SampleDemographic and Clinical Characteristics of the Overall Sample

Table 1:

Demographic and Clinical Characteristics of the Overall Sample

Predictors of Patient Satisfaction With the Hospital Encounter

The multiple logistic regression results (areas under the curve ranged from 0.5862 to 0.6534), given fixed values of all other variables in the model, for each of the 10 HCAHPS domain outcome measures are shown in Tables A, B, C, D, E, F, G, and H (available in the online version of the article). Increased LOS in the hospital was associated with lower predicted odds of being completely satisfied with the responsiveness of the hospital staff (OR, 0.864; 95% CI, 0.768–0.960; P=.0101; Figure 1). However, obese patients (BMI ≥30 kg/m2) were more likely to be completely satisfied with the responsiveness of the hospital staff (OR, 1.451; 95% CI, 1.048–2.015; P=.0271) than nonobese patients (BMI <30 kg/m2). The predicted odds of being completely satisfied regarding communication with nurses were also greater for patients who were obese (OR, 1.398; 95% CI, 0.991–1.981; P=.0601) and who had arthroplasty (OR, 1.477; 95% CI, 1.019–2.148; P=.0423), but were lower if the care provider was younger than the patient (OR, 0.441; 95% CI, 0.259–0.739; P=.0024). The odds of being completely satisfied regarding communication with physicians, discharge information, and responsiveness of the hospital staff were also significantly lower if the care provider was younger than the patient (OR, 0.527, 0.339, and 0.595, respectively; P=.0473). Male patients were more likely to be completely satisfied regarding communication about medicines (OR, 1.694; 95% CI, 1.130–2.554; P=.0126), care transition (OR, 1.489; 95% CI, 1.067–2.079; P=.0208), and cleanliness of the hospital environment (OR, 2.120; 95% CI, 1.421–3.202; P=.0003) than female patients. Married patients were less likely to be completely satisfied with the cleanliness of the hospital environment (OR, 0.501; 95% CI, 0.314–0.777; P=.0027) than nonmarried patients.

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication with Nurses

Table A:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication with Nurses

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication with Doctors

Table B:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication with Doctors

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Responsiveness of Hospital Staff

Table C:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Responsiveness of Hospital Staff

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication about Medicines

Table D:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication about Medicines

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Discharge Information

Table E:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Discharge Information

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Care Transition

Table F:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Care Transition

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Cleanliness of Hospital Environment

Table G:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Cleanliness of Hospital Environment

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Recommend the Hospital

Table H:

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Recommend the Hospital

Plot of the predicted probability of patient satisfaction with the responsiveness of the hospital staff as a function of length of hospital stay, with 95% confidence intervals for the regression surface, from the hospital encounters of 646 patients. Multiple logistic regression was implemented to estimate the predicted probabilities of patient satisfaction from length of hospital stay, given fixed values of all other variables in the model.

Figure 1:

Plot of the predicted probability of patient satisfaction with the responsiveness of the hospital staff as a function of length of hospital stay, with 95% confidence intervals for the regression surface, from the hospital encounters of 646 patients. Multiple logistic regression was implemented to estimate the predicted probabilities of patient satisfaction from length of hospital stay, given fixed values of all other variables in the model.

The predicted odds of being completely satisfied with the care transition were also significantly greater for Medicare-insured patients than for privately insured patients (OR, 1.748; 95% CI, 1.184–2.597; P=.0058) and for obese patients (OR, 1.416; 95% CI, 1.021–1.967; P=.0395) than for nonobese patients, but were lower if the number of patient consults was 2 to 5 vs 0 (OR, 0.573; 95% CI, 0.329–0.988; P=.0495).

Increased LOS in the hospital (OR, 0.920; 95% CI, 0.847–0.998; P=.0498; Figure 2) and the care provider being younger than the patient (OR, 0.547; 95% CI, 0.296–0.989; P=.0496) were significantly associated with lower odds of the patient recommending the hospital (with complete satisfaction). At a trend level, the odds of recommending the hospital (with complete satisfaction) were greater for male patients (OR, 1.476; 95% CI, 0.968–2.273; P=.0737) than for female patients, for White patients (OR, 1.690; 95% CI, 0.947–3.020; P=.0763) than for non-White patients, and as patient age increased (OR, 1.025; 95% CI, 0.998–1.053; P=.0716).

Plot of the predicted probability of patient satisfaction with recommending the hospital as a function of length of hospital stay, with 95% confidence intervals for the regression surface, from the hospital encounters of 646 patients. Multiple logistic regression was implemented to estimate the predicted probabilities of patient satisfaction from length of hospital stay, given fixed values of all other variables in the model.

Figure 2:

Plot of the predicted probability of patient satisfaction with recommending the hospital as a function of length of hospital stay, with 95% confidence intervals for the regression surface, from the hospital encounters of 646 patients. Multiple logistic regression was implemented to estimate the predicted probabilities of patient satisfaction from length of hospital stay, given fixed values of all other variables in the model.

No significant predictors of patient satisfaction emerged from the logistic regression analysis for quietness of the hospital environment or for the global hospital rating.

Discussion

This study indicates that patient factors are predictive of patient satisfaction within various domains measured by the HCAHPS. Certain patient characteristics may be important to consider when determining overall patient satisfaction. The authors evaluated 14 characteristic variables, including LOS, and their effects on each of the 10 domains. Among the variables, the ones that predicted the odds of patient satisfaction included patient BMI, patient sex, patient race, patient marital status, provider age, insurance provider, number of consults, and LOS.

Some studies have found that LOS may be associated with certain patient satisfaction.21–24 With Medicare's bundling of payments for procedures such as total hip arthroplasty (THA), LOS also affects hospital efficiency and profitability.7 Within current bundled payment programs, hospitals are paid a fixed sum for all care within 90 days of surgery, and postoperative care represents up to 40% of costs of surgeries like THA.7 Current literature on the effect of LOS on satisfaction within individual domains of the HCAHPS is sparse. This study found that patients with longer LOS have lower odds of being completely satisfied and recommending the hospital. This may be related to patient frustration with expectations of care and care provider fatigue. Length of stay is modifiable. Surgeries scheduled for earlier in the week are associated with shorter LOS7 and decreased 90-day read-mission.25 Surgeons can improve satisfaction scores and hospital costs by planning surgeries on earlier days of the week, allowing patients to receive rehabilitation and thorough discharge planning when the hospital is at full staff.

Patient sex plays a role in patient satisfaction.26 The current authors have found that males are more likely to be satisfied with communication about medications, care transition, and cleanliness. Males were also more likely to recommend the hospital. Other studies have also found male patients to be more likely to report satisfaction with the clinical encounter.23,27,28 Although patient sex is nonmodifiable, surgeons may improve satisfaction in domains such as care transition by planning postoperative rehabilitation that is specific to patient factors such as sex, education level, and socioeconomic barriers to care.

In this study, Medicare as primary payer was associated with better satisfaction with care transition. Studies have found that self-paying and Medicaid patients have higher HCAHPS scores.29,30 Interestingly, lower socioeconomic status has been shown to be associated with higher satisfaction24 but poorer outcomes.31 Higher levels of education have also been found to be associated with lower satisfaction.32 However, the current authors also found that a higher number of consults was associated with decreased odds of satisfaction with care transition. Care transition scores can be improved by implementing a multidisciplinary approach with all care teams involved in discharge planning. Care transition scores can also be improved by sensitivity to patient factors that may hinder postdischarge care.

This study found that older patients were more likely to recommend the hospital. This finding is consistent with that of Hall and Dornan,33 who found that younger patients tend to be less satisfied with the care they received, and Bourne et al,34 who found that older patients were more satisfied with their care. Older patients simply might have more experience with hospital systems and thus have more realistic expectations.

Patients who had a care provider younger than themselves reported decreased satisfaction regarding communication with physicians and nurses, discharge information, and responsiveness of the hospital staff, and they were less likely to recommend the hospital to others. Patients may associate age with competence. Young providers may face an uphill battle early in their careers. Better communication skills may also come with experience. As physicians learn how to diagnose their patients, effective communication skills should not be ignored. These results highlight the need for studies that evaluate which communication techniques are most effective for various age groups. By tailoring communication styles to each patient, physicians can improve satisfaction. Better communication will also increase comprehension to improve outcomes.

Although the authors' findings regarding the relationship between patient age, insurance provider, patient sex, ethnicity, and LOS were supported by other studies, there exists a paucity of extant research that evaluates patient BMI as a factor of patient satisfaction with care received. The authors found that obese patients (BMI ≥30 kg/m2) were more likely to be satisfied with the responsiveness of the hospital staff and care transition. This belies their expectations. The authors expected patients with higher BMI to be associated with lower patient satisfaction because of, in part, increased pain and difficulty with mobility and higher incidence of postoperative morbidities.35 The authors believe further research on BMI and its association with patient satisfaction is warranted.

This study's limitations included the retrospective design. Data were collected at a single institution whose patient population may not represent the populations of other hospitals. A larger sample can allow patients to be evaluated by type of orthopedic subspecialty, which may play a role in patient satisfaction. However, the sample size was limited by the number of patients who completed the survey at the authors' institution. Data from patients who did not complete the survey are not available, and the number of patients who did not complete is not available. However, poor response rates to the HCAHPS survey is a weakness inherent to the survey and would be a limitation of any study of HCAHPS. Tyser et al36 found response rates to surveys such as the Press Ganey, the precursor to HCAHPS, are poor and affected by patient characteristics. Despite these limitations, the HCAHPS survey continues to be used in assessing hospital quality and reimbursement. Patients undergoing more invasive procedures with longer recovery times may have skewed satisfaction, which also represents a weakness of the HCAHPS survey. A multicenter study would also increase the diversity of the patient population, although the authors' center is a tertiary referral center serving a major US city.

To correctly compute total scores for each HCAHPS domain, which is a function of <n> items per domain, missing <n> items had to be either excluded for a given patient or imputed (so that a total score could be calculated pursuant to the scoring methodology of CMS). Because only a small proportion of values was missing (<10%), the authors chose to impute missing items as opposed to exclude a given patient. Missing values (with an assumed arbitrary missing pattern) for the classification variables and for the continuous variables were imputed via 500 burn-in iterations (samples) using fully conditional specification along with the discriminant method (for the classification variables) and the predictive mean matching method (for continuous variables), which preserved the a priori (before imputation) distribution of each item and variable. That is, the imputation did not lead to a skewed item or variable distribution (it retained the basic distribution that was already in place before the imputation). Despite the limitations of this study, it allows physicians and hospital systems to understand how individual domains are affected by patient factors. Furthermore, additional studies evaluating the factors that affect patient satisfaction can use these findings as a starting point in efforts toward better patient care and outcomes.

Conclusion

The authors have sought to evaluate the effect of hospital LOS on each of the 10 domains of HCAHPS patient satisfaction. Longer LOS led to lower patient satisfaction and a lower likelihood of recommending the hospital. The authors have also found that other factors such as provider younger than patient, lower BMI, female sex, non-Medicare (private) insurance, and a higher number of consults were associated with lower patient satisfaction in various domains. Patient outcomes are no longer the only metrics used to measure care providers' abilities. As employers, insurance companies, and patients place more emphasis on care provider and hospital ratings, understanding the various factors that affect patient satisfaction and how to improve patient satisfaction are paramount. With heterogeneous patient populations, care providers must learn to adjust communication styles, discharge planning, and surgery scheduling. By improving patient satisfaction on individual domains of the HCAHPS survey, care providers can improve their overall scores, which can lead to better patient outcomes.

References

  1. HCAHPS fact sheet November 2017. www.hcahpsonline.org/globalassets/hcahps/facts/hcahps_fact_sheet_november_2017.pdf. Accessed January 13, 2019.
  2. Giordano LA, Elliott MN, Goldstein E, Lehrman WG, Spencer PA. Development, implementation, and public reporting of the HCAHPS survey. Med Care Res Rev. 2010;67(1):27–37. doi:10.1177/1077558709341065 [CrossRef] PMID:19638641
  3. Shulman B, Hutzler L, Karia R, Bosco J. The effect of length of stay and discharge disposition on hospital consumer assessment of healthcare providers and systems scores in orthopaedic patients. J Am Acad Orthop Surg. 2019;27(9):e418–e422. doi:10.5435/jaaos-d-17-00552 [CrossRef] PMID:30379757
  4. Jha AK, Orav EJ, Zheng J, Epstein AM. Patients' perception of hospital care in the United States. N Engl J Med. 2008;359(18):1921–1931. doi:10.1056/NEJMsa0804116 [CrossRef] PMID:18971493
  5. Tsai TC, Orav EJ, Jha AK. Patient satisfaction and quality of surgical care in US hospitals. Ann Surg. 2015;261(1):2–8. doi:10.1097/SLA.0000000000000765 [CrossRef] PMID:24887985
  6. Sacks GD, Lawson EH, Dawes AJ, et al. Relationship between hospital performance on a patient satisfaction survey and surgical quality. JAMA Surg. 2015;150(9):858–864. doi:10.1001/jamasurg.2015.1108 [CrossRef] PMID:26108091
  7. Keswani A, Beck C, Meier KM, Fields A, Bronson MJ, Moucha CS. Day of surgery and surgical start time affect hospital length of stay after total hip arthroplasty. J Arthroplasty. 2016;31(11):2426–2431. doi:10.1016/j.arth.2016.04.013 [CrossRef] PMID:27491449
  8. Jeon CY, Neidell M, Jia H, Sinisi M, Larson E. On the role of length of stay in healthcare-associated bloodstream infection. Infect Control Hosp Epidemiol. 2012;33(12):1213–1218. doi:10.1086/668422 [CrossRef] PMID:23143358
  9. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–383. doi:10.1016/0021-9681(87)90171-8 [CrossRef] PMID:3558716
  10. Technical notes for HCAHPS star ratings: revised for October 2018 public reporting. https://www.hcahpsonline.org/globalassets/hcahps/star-ratings/tech-notes/2018-10_star-ratings_tech-notes.pdf. Accessed January 13, 2019.
  11. Martin L, Presson AP, Zhang C, Ray D, Finlayson S, Glasgow R. Association between surgical patient satisfaction and nonmodifiable factors. J Surg Res. 2017;214:247–253. doi:10.1016/j.jss.2017.03.029 [CrossRef] PMID:28624052
  12. Lubitz RMThe Society of General Internal Medicine Education Committee. Guidelines for promotion of clinician-educators. J Gen Intern Med. 1997;12(S2)(suppl 2):S71–S78. doi:10.1046/j.1525-1497.12.s2.10.x [CrossRef] PMID:9127247
  13. 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 [CrossRef] PMID:27004489
  14. Abtahi AM, Presson AP, Zhang C, Saltzman CL, Tyser AR. Association between orthopaedic outpatient satisfaction and non-modifiable patient factors. J Bone Joint Surg Am. 2015;97(13):1041–1048. doi:10.2106/JBJS.N.00950 [CrossRef] PMID:26135070
  15. Sun BC, Adams J, Orav EJ, Rucker DW, Brennan TA, Burstin HR. Determinants of patient satisfaction and willingness to return with emergency care. Ann Emerg Med. 2000;35(5):426–434. doi:10.1016/S0196-0644(00)70003-5 [CrossRef] PMID:10783404
  16. Mira JJ, Tomás O, Virtudes-Pérez M, Nebot C, Rodríguez-Marín J. Predictors of patient satisfaction in surgery. Surgery. 2009;145(5):536–541. doi:10.1016/j.surg.2009.01.012 [CrossRef] PMID:19375613
  17. Schmocker RK, Cherney Stafford LM, Siy AB, Leverson GE, Winslow ER. Understanding the determinants of patient satisfaction with surgical care using the Consumer Assessment of Healthcare Providers and Systems surgical care survey (SCAHPS). Surgery. 2015;158(6):1724–1733. doi:10.1016/j.surg.2015.06.018 [CrossRef] PMID:26195107
  18. Weiss E. Knee osteoarthritis, body mass index and pain: data from the Osteoarthritis Initiative. Rheumatology (Oxford). 2014;53(11):2095–2099. doi:10.1093/rheumatology/keu244 [CrossRef] PMID:24939675
  19. Christensen R, Henriksen M, Leeds AR, et al. Effect of weight maintenance on symptoms of knee osteoarthritis in obese patients: a twelve-month randomized controlled trial. Arthritis Care Res (Hoboken).2015;67(5):640–650. doi:10.1002/acr.22504 [CrossRef] PMID:25370359
  20. van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Stat Methods Med Res. 2007;16(3):219–242. doi:10.1177/0962280206074463 [CrossRef] PMID:17621469
  21. Kemp KA, Santana MJ, Southern DA, Mc-Cormack B, Quan H. Association of inpatient hospital experience with patient safety indicators: a cross-sectional, Canadian study. BMJ Open. 2016;6(7):e011242. doi:10.1136/bmjopen-2016-011242 [CrossRef] PMID:27371554
  22. Mistry JB, Chughtai M, Elmallah RK, et al. What influences how patients rate their hospital after total hip arthroplasty?J Arthroplasty. 2016;31(11):2422–2425. doi:10.1016/j.arth.2016.03.060 [CrossRef]
  23. Peres-da-Silva A, Kleeman LT, Wellman SS, et al. What factors drive inpatient satisfaction after knee arthroplasty?J Arthroplasty.2017;32(6):1769–1772. doi:10.1016/j.arth.2017.01.036 [CrossRef]
  24. Maratt JD, Lee YY, Lyman S, Westrich GH. Predictors of satisfaction following total knee arthroplasty. J Arthroplasty. 2015;30(7):1142–1145. doi:10.1016/j.arth.2015.01.039 [CrossRef]
  25. Boylan MR, Perfetti DC, Naziri Q, Maheshwari AV, Paulino CB, Mont MA. Is day of surgery associated with adverse clinical and economic outcomes following primary total knee arthroplasty?J Arthroplasty. 2017;32(8):2339–2346. doi:10.1016/j.arth.2017.03.029 [CrossRef] PMID:28413140
  26. Elliott MN, Lehrman WG, Beckett MK, Goldstein E, Hambarsoomian K, Giordano LA. Gender differences in patients' perceptions of inpatient care. Health Serv Res. 2012;47(4):1482–1501. doi:10.1016/10.1111/j.1475-6773.2012.01389.x [CrossRef]
  27. Eftekhary N, Feng JE, Anoushiravani AA, Schwarzkopf R, Vigdorchik JM, Long WJ. Revision versus primary hospital consumer assessment of healthcare providers and systems scores in total joint arthroplasty. J Arthroplasty. 2019;34(7S):S84–S90. doi:10.1016/j.arth.2018.11.019 [CrossRef] PMID:30545652
  28. Levinson W, Frankel RM, Roter D, Drum M. How much do surgeons like their patients?Patient Educ Couns.2006;61(3):429–434. doi:10.1016/j.pec.2005.05.009 [CrossRef] PMID:16024213
  29. Danforth RM, Flanagan ME, Brewster BD, Brand EW, Frankel RM, Pitt HA. Surgical inpatient satisfaction: what are the real drivers?J Surg Res. 2014;186(2):531. doi:10.1016/j.jss.2013.11.388 [CrossRef]
  30. Liao L, Jollis JG, DeLong ER, Peterson ED, Morris KG, Mark DB. Impact of an interactive video on decision making of patients with ischemic heart disease. J Gen Intern Med. 1996;11(6):373–376. doi:10.1007/BF02600051 [CrossRef] PMID:8803746
  31. Judge A, Arden NK, Cooper C, et al. Predictors of outcomes of total knee replacement surgery. Rheumatology (Oxford). 2012;51(10):1804–1813. doi:10.1093/rheumatology/kes075 [CrossRef] PMID:22532699
  32. Li L, Lee NJ, Glicksberg BS, Radbill BD, Dudley JT. Data-driven identification of risk factors of patient satisfaction at a large urban academic medical center. PLoS One. 2016;11(5):e0156076. doi:10.1371/journal.pone.0156076 [CrossRef] PMID:27228056
  33. Hall JA, Dornan MC. Patient sociodemographic characteristics as predictors of satisfaction with medical care: a meta-analysis. Soc Sci Med. 1990;30(7):811–818. doi:10.1016/0277-9536(90)90205-7 [CrossRef] PMID:2138357
  34. Bourne RB, Chesworth BM, Davis AM, Mahomed NN, Charron KD. Patient satisfaction after total knee arthroplasty: who is satisfied and who is not?Clin Orthop Relat Res.2010;468(1):57–63. doi:10.1007/s11999-009-1119-9 [CrossRef] PMID:19844772
  35. Escallon J. Postoperative complications in obese and non-obese patients. World J Surg. 2007;31:561. doi:10.1007/s00268-006-0422-9 [CrossRef]
  36. Tyser AR, Abtahi AM, McFadden M, Presson AP. Evidence of non-response bias in the Press-Ganey patient satisfaction survey. BMC Health Serv Res. 2016;16(a):350. doi:10.1186/s12913-016-1595-z [CrossRef] PMID:27488567

Demographic and Clinical Characteristics of the Overall Sample

CharacteristicValue
Patient demographics
  Age, mean (SD), y67.14 (11.96)
  Male, No.290 (44.89%)
  White, non-Hispanic, No.511 (79.10%)
  Married, No.457 (70.74%)
Care provider demographics, No.
  Same race as patient319 (49.38%)
  Younger than patient478 (73.99%)
Patient factors
  BMI
    Mean (SD), kg/m229.54 (6.14)
    <30 kg/m2, No.369 (57.12%)
    ≥30 kg/m2, No.277 (42.88%)
  LOS in hospital, mean (SD), d2.57 (2.55)
  Consults, No.
    0415 (64.24%)
    1136 (21.05%)
    2–595 (14.71%)
  Admission due to previous complication, No.72 (11.15%)
  Medicare health insurance status, No.311 (48.14%)
  Private health insurance status, No.335 (51.86%)
Patient comorbidities
  Charlson Comorbidity Index, mean (SD)3.13 (1.97)
  Allergies, mean (SD), No.1.84 (3.25)
HCAHPS patient satisfaction scoresa
  Composite measures
    Communication with nurses, mean (SD)91.39 (16.08)
      Completely satisfied (HCAHPS score=100), No.431 (66.72%)
    Communication with physicians, mean (SD)95.61 (11.35)
      Completely satisfied (HCAHPS score=100), No.523 (80.96%)
    Responsiveness of hospital staff, mean (SD)84.28 (21.15)
      Completely satisfied (HCAHPS score=100), No.348 (53.87%)
    Communication about medicines, mean (SD)79.58 (26.53)
      Completely satisfied (HCAHPS score=100), No.218 (50.70%)
    Discharge information, mean (SD)93.11 (18.96)
      Completely satisfied (HCAHPS score=100), No.565 (87.46%)
    Care transition, mean (SD)84.87 (17.54)
      Completely satisfied (HCAHPS score=100), No.301 (46.59%)
  Individual items
    Cleanliness of hospital environment, mean (SD)88.63 (23.37)
      Completely satisfied (HCAHPS score=100), No.493 (76.32%)
    Quietness of hospital environment, mean (SD)88.91 (19.93)
      Completely satisfied (HCAHPS score=100), No.468 (72.45%)
  Global items
    Global hospital rating, mean (SD)89.34 (17.63)
      Completely satisfied (HCAHPS score=100), No.345 (53.41%)
    Recommend the hospital, mean (SD)91.62 (19.43)
      Completely satisfied (HCAHPS score=100), No.523 (80.96%)
  Care provider subspecialty, No.
    Arthroplasty349 (54.02%)
    Spine244 (37.77%)
    Trauma26 (4.02%)
    Foot ankle15 (2.32%)
    Sports5 (0.77%)
    Infection5 (0.77%)
    Hand2 (0.31%)

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication with Nurses

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years1.0140.991 to 1.0370.2458
  Sex (Male vs. Female)1.1070.779 to 1.5780.5752
  Race (White vs. Non-White)1.1880.732 to 1.9260.4891
  Marital Status (Married vs. Not Married)1.0790.741 to 1.5630.6923
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.4410.259 to 0.7390.0024
  Same Race as Patient (Yes vs. No)0.9650.623 to 1.4880.8742
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 10.8010.508 to 1.2670.3447
  Number of Consults, 2–50.6320.368 to 1.0890.1014
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.3980.991 to 1.9810.0601
  Admission due to previous complication (Yes vs. No)1.1710.676 to 2.0900.5867
  Health Insurance (Medicare vs. Private)1.2710.850 to 1.9050.2490
  Length of Stay in Hospital, days0.9560.880 to 1.0220.2388
Patient Comorbidities
  Charlson Comorbidity Index Score0.9550.853 to 1.0730.4416
  Number of Allergies1.0120.961 to 1.0800.6920
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)1.4771.019 to 2.1480.0423

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication with Doctors

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years0.9930.965 to 1.0210.6406
  Sex (Male vs. Female)1.0350.681 to 1.5790.8725
  Race (White vs. Non-White)1.1750.670 to 2.0450.5718
  Marital Status (Married vs. Not Married)1.1930.766 to 1.8380.4299
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.5270.276 to 0.9800.0473
  Same Race as Patient (Yes vs. No)0.8620.513 to 1.4380.5721
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 10.7630.454 to 1.2990.3138
  Number of Consults, 2–50.6300.343 to 1.1790.1452
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.2380.822 to 1.8780.3127
  Admission due to previous complication (Yes vs. No)0.7710.428 to 1.4450.4054
  Health Insurance (Medicare vs. Private)1.2200.758 to 1.9670.4146
  Length of Stay in Hospital, days0.9810.913 to 1.0580.6235
Patient Comorbidities
  Charlson Comorbidity Index Score1.1300.976 to 1.3280.1176
  Number of Allergies0.9640.909 to 1.0170.2115
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)0.7160.455 to 1.1170.1442

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Responsiveness of Hospital Staff

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years1.0150.993 to 1.3710.1837
  Sex (Male vs. Female)1.2580.904 to 1.7540.1791
  Race (White vs. Non-White)0.9590.604 to 1.5240.8622
  Marital Status (Married vs. Not Married)1.0620.744 to 1.5140.7420
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.5950.366 to 0.9580.0360
  Same Race as Patient (Yes vs. No)1.0130.669 to 1.5310.9512
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 11.0510.683 to 1.6240.8231
  Number of Consults, 2–50.8400.491 to 1.4390.5293
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.4511.048 to 2.0150.0271
  Admission due to previous complication (Yes vs. No)1.2310.727 to 2.1100.4487
  Health Insurance (Medicare vs. Private)1.0510.717 to 1.5420.8016
  Length of Stay in Hospital, days0.8640.768 to 0.9600.0101
Patient Comorbidities
  Charlson Comorbidity Index Score0.9870.883 to 1.1020.8161
  Number of Allergies1.0040.956 to 1.0570.8714
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)1.0460.734 to 1.4930.8048

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Communication about Medicines

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years1.0070.981 to 1.0340.6046
  Sex (Male vs. Female)1.6941.130 to 2.5540.0126
  Race (White vs. Non-White)1.0350.606 to 1.7710.9010
  Marital Status (Married vs. Not Married)0.7480.487 to 1.1440.1900
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.7220.398 to 1.2990.2874
  Same Race as Patient (Yes vs. No)1.2020.738 to 1.9590.4672
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 11.1070.675 to 1.8220.6938
  Number of Consults, 2–50.7340.408 to 1.3110.3074
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.1710.790 to 1.7390.4387
  Admission due to previous complication (Yes vs. No)1.1330.575 to 2.2530.7252
  Health Insurance (Medicare vs. Private)1.0510.717 to 1.5420.8016
  Length of Stay in Hospital, days0.9740.886 to 1.0460.5315
Patient Comorbidities
  Charlson Comorbidity Index Score0.9930.871 to 1.1320.9150
  Number of Allergies1.0270.972 to 1.1020.4124
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)0.7580.489 to 1.1740.2231

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Discharge Information

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years1.0120.979 to 1.0450.4663
  Sex (Male vs. Female)1.5590.939 to 2.6190.0837
  Race (White vs. Non-White)1.1200.533 to 2.3040.7569
  Marital Status (Married vs. Not Married)1.5210.907 to 2.5140.1026
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.3390.148 to 0.7330.0069
  Same Race as Patient (Yes vs. No)0.6150.326 to 1.1350.1216
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 10.9840.523 to 1.9200.9613
  Number of Consults, 2–51.2200.560 to 2.8730.6273
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.3030.801 to 2.1500.2871
  Admission due to previous complication (Yes vs. No)0.8090.392 to 1.8300.5847
  Health Insurance (Medicare vs. Private)1.0410.592 to 1.8280.8863
  Length of Stay in Hospital, days0.9590.889 to 1.0580.3425
Patient Comorbidities
  Charlson Comorbidity Index Score0.9900.838 to 1.1920.9073
  Number of Allergies1.0890.983 to 1.2450.1386
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)1.0950.652 to 1.8400.7275

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Care Transition

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years0.9870.966 to 1.0090.2587
  Sex (Male vs. Female)1.4891.067 to 2.0790.0208
  Race (White vs. Non-White)1.2350.774 to 1.9780.3828
  Marital Status (Married vs. Not Married)1.2580.878 to 1.8080.2173
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.6870.424 to 1.1070.1284
  Same Race as Patient (Yes vs. No)0.9250.610 to 1.4000.7143
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 11.1080.719 to 1.7110.6456
  Number of Consults, 2–50.5730.329 to 0.9880.0495
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.4161.021 to 1.9670.0395
  Admission due to previous complication (Yes vs. No)1.4270.841 to 2.4380.1959
  Health Insurance (Medicare vs. Private)1.7481.184 to 2.5970.0058
  Length of Stay in Hospital, days0.9690.875 to 1.0440.4738
Patient Comorbidities
  Charlson Comorbidity Index Score0.9660.862 to 1.0800.5531
  Number of Allergies0.9750.920 to 1.0250.3695
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)1.0940.767 to 1.5640.6240

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Cleanliness of Hospital Environment

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years1.0110.985 to 1.0360.4222
  Sex (Male vs. Female)2.1201.421 to 3.2020.0003
  Race (White vs. Non-White)1.2200.713 to 2.0790.4682
  Marital Status (Married vs. Not Married)0.5010.314 to 0.7770.0027
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.8270.469 to 1.4400.5094
  Same Race as Patient (Yes vs. No)1.2930.795 to 2.0990.3016
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 11.3190.785 to 2.2620.3063
  Number of Consults, 2–50.8010.451 to 1.4440.4596
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.1940.817 to 1.7540.3661
  Admission due to previous complication (Yes vs. No)0.7490.421 to 1.3720.3400
  Health Insurance (Medicare vs. Private)0.9350.594 to 1.4700.7731
  Length of Stay in Hospital, days0.9770.909 to 1.0510.5376
Patient Comorbidities
  Charlson Comorbidity Index Score0.9550.840 to 1.0920.4926
  Number of Allergies1.0220.965 to 1.1040.5256
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)1.2180.806 to 1.8450.3524

Odds Ratios from the Multiple Logistic Regression for Predictors of Patient Satisfaction regarding Recommend the Hospital

Predictor VariablesAdjusted Odds Ratio95% CI for Adjusted Odds RatioP-value
Patient Demographics
  Age, years1.0250.998 to 1.0530.0716
  Sex (Male vs. Female)1.4760.968 to 2.2730.0737
  Race (White vs. Non-White)1.6900.947 to 3.0200.0763
  Marital Status (Married vs. Not Married)0.9060.570 to 1.4160.6716
Care Provider Demographics
  Provider is younger in age than patient (Yes vs. No)0.5470.296 to 0.9890.0496
  Same Race as Patient (Yes vs. No)0.7580.439 to 1.2910.3125
Patient Factors
  Number of Consults, 0 (reference group)̶̶̶
  Number of Consults, 10.9800.574 to 1.7060.9416
  Number of Consults, 2–51.2250.634 to 2.4740.5596
  BMI group (≥30 kg/m2 vs. <30 kg/m2)1.1390.757 to 1.7270.5352
  Admission due to previous complication (Yes vs. No)1.3970.707 to 3.0230.3642
  Health Insurance (Medicare vs. Private)1.0420.644 to 1.6850.8678
  Length of Stay in Hospital, days0.9200.847 to 0.9980.0498
Patient Comorbidities
  Charlson Comorbidity Index Score0.9650.842 to 1.1190.6233
  Number of Allergies0.9760.926 to 1.0370.4064
Care Provider Subspecialty
  Arthroplasty (Yes vs. No)1.2710.819 to 1.9730.2861
Authors

The authors are from the Department of Orthopaedic Surgery (WD, JW) and the Department of Population and Data Sciences (PAN), University of Texas Southwestern Medical School, Dallas, Texas.

The authors have no relevant financial relationships to disclose.

Correspondence should be addressed to: Joel Wells, MD, MPH, Department of Orthopaedic Surgery, University of Texas Southwestern Medical School, 1801 Inwood Rd, Dallas, TX 75390 ( joel.wells@UTSouthwestern.edu).

Received: August 13, 2019
Accepted: September 24, 2019
Posted Online: September 22, 2020

10.3928/01477447-20200910-02

Sign up to receive

Journal E-contents