Preoperative predictors of two-year satisfaction in hand and wrist surgery patients

Abstract

Purpose

As stakeholders seek to improve patient outcomes while maintaining cost-effectiveness in an increasingly expensive healthcare system, metrics such as patient satisfaction are becoming more important. This present study sought to identify factors associated with and independently predictive of better surgical satisfaction two years following hand and wrist surgery.

Methods

Patients undergoing hand and wrist surgery at an urban outpatient institution were enrolled preoperatively into a surgical registry and assessed two years postoperatively. Patient satisfaction with surgery was measured at two years postoperatively with the Surgical Satisfaction Questionnaire (SSQ-8). Bivariate analysis determined associations between postoperative satisfaction and patient demographics, injury specifiers, medical history, and multiple patient-reported outcomes (PROs). Multivariable analysis determined independent predictors of two-year postoperative satisfaction following hand and wrist surgery.

Results

Better surgical satisfaction was associated with having never smoked, no preoperative opioid use, lack of an accompanying legal claim, lack of a workers compensation claim, no clinical history of depression/anxiety, less comorbidities, and higher preoperative expectations.

Various PROs relating to function, pain, activity, and general health at both baseline and two years demonstrated associations with postoperative satisfaction. Multivariable analysis confirmed that never smoking, lack of a legal claim, and better preoperative Brief Michigan Hand Questionnaire scores were independently predictive of better surgical satisfaction two years following hand and wrist surgery.

Conclusion

At two years following hand and wrist surgery, better patient satisfaction was best predicted by never smoking, no related legal claim, and better baseline Brief Michigan Hand Questionnaire scores.

Level of evidence

III.

1. Introduction

Tracking patient outcomes following interactions with the healthcare system is essential for providing and improving patient-centered care. Patient satisfaction is a complex consideration used expansively to determine reimbursement rates, evaluate provider quality of care, improve patient experiences, and help guide patient decision making.¹ Although satisfaction can have different meanings to different patients, it can ultimately dictate a patient's willingness to maintain follow-up.2, 3, 4, 5 Satisfaction is predominantly measured nationwide by the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey, which is tied into Medicare's value-based incentive payments via the Affordable Care Act.⁶ Satisfaction is also measured through third-party surveys such as Press Ganey as well as institution and specialty-specific appraisals.² However, these overarching satisfaction scores may not be appropriate within the hand and wrist surgery population given an overall low response rate and evidence finding a lack of correlation with outcomes in other surgical interventions.7, 8, 9, 10, 11

It is crucial to discern underpinnings of surgical satisfaction within hand and wrist surgery patients, especially given the high burden of hand and wrist pathology and the continued rise of healthcare costs.12, 13, 14, 15, 16 Despite the heterogenous existing data, improvement of pain, hand and wrist symptoms, and aesthetics all have been shown to have the strongest correlation to satisfaction, while other PROs and patient history factors were found to have inconsistent correlations with satisfaction across various studies.¹⁷ Many of these PROs, however, are measurements specific only to individual procedures within hand and wrist surgery.17, 18, 19, 20, 21, 22 Postoperative hand appearance and various metrics of psychological comorbidities have demonstrated associations with satisfaction, while external factors such as navigation of finances or straightforward rehabilitation have been highlighted in open-ended patient surveys.^18,19,23, 24, 25, 26 Factors important for postoperative satisfaction were distinct from what drove satisfaction in clinic or the preoperative period preparing for the procedure.27, 28, 29, 30 These studies characterize the associations of many postoperative measures as they relate to satisfaction, however, there is a gap assessing preoperative measures and patient demographic factors in hand and wrist surgery patients’ postoperative satisfaction. The importance of considering hand and wrist surgery in isolation is further highlighted by the range of predictors for satisfaction following surgery in other surgical subspecialties such as hip and knee arthroplasty, shoulder surgery, and spine surgery.³¹

The goal of this study was to evaluate predictors of patient satisfaction two years following hand and wrist surgery. We hypothesize that better surgical satisfaction at two years postoperatively will be predicted by higher preoperative expectations, higher degree of education, absence of depression and anxiety, and better general health.

2. Methods

2.1. Cohort selection

Between July 2015 and July 2018, prospective survey data was voluntarily collected with informed consent from patients undergoing surgery at a single outpatient institution under the care of four surgeons. A retrospective analysis of the survey data for patients older than 16 who had hand or wrist surgery cataloged in the Institutional Review Board (IRB) approved Maryland Orthopaedic Registry was performed.³² Patients completed baseline surveys within one week of surgery and follow-up surveys at the two-year timepoint. Patient data was de-identified by associating their medical record number to a unique participant identification number within the registry.

2.2. Survey data collection

Baseline survey fields included demographic information (i.e. smoking status – current, former or never smoker, frequency of alcohol consumption, income, education level, etc.), medical and surgical history, and various patient-reported outcomes (PROs). These PROs included Patient-Reported Outcomes Measurement Information System (PROMIS®) computer adaptive testing (CAT) domains of Physical Function (PF), Pain Interference (PI), Fatigue, Social Satisfaction (SS), Anxiety, and Depression.³³ Patient's level of activity and function prior to surgery was captured by the Tegner Activity Scale and the Marx Activity Rating Scale - Upper Extremity.^34,35 Hand function and ability to perform tasks of daily living was assessed with the Brief Michigan Hand Questionnaire (BMHQ), a validated and quicker version of the original Michigan Hand Questionnaire.³⁶ Pain at the operative site was assessed with the Numeric Pain Scale (NPS).³⁷ Preoperative expectations of the surgery were recorded using the Musculoskeletal Outcomes Data Evaluation and Management System (MODEMS) expectations domain.³⁸ Survey data was collected from patients via the Research Electronic Data Capture (REDCap™) data collection system.³⁹ A detailed chart review was performed to obtain body mass index (BMI) measurements, use of preoperative opioid medications, American Society of Anesthesiologists (ASA) score, and burden of comorbidities as rated by Charlson Comorbidity Index (CCI), in which a higher score represents greater health issues.⁴⁰ Operative notes were individually reviewed to verify that Common Procedural Terminology (CPT) code(s) appropriate for hand and wrist surgery were included.

Surgical satisfaction was assessed at two-years postoperatively using the Surgical Satisfaction Questionnaire (SSQ-8).⁴¹ This questionnaire gauges patient's satisfaction with pain control, time to return to various stages of activities, and overall results; culminating with questions of willingness to repeat the surgery if necessary and likelihood to recommend the surgery to another person. Scores of the SSQ-8 were determined by each question being rated from 1 (“very satisfied) to 5 (“very unsatisfied”) then aggregated over the eight questions before being normalized to 100, with 100 being the highest amount of overall satisfaction.

To compare questionnaire results with baseline answers, the two-year follow-up surveys also included the same six PROMIS domains, postoperative Tegner and Marx activity scales, postoperative NPS at operative site, and postoperative BMHQ questionnaire. The level to which a patient's expectations regarding surgery were met was assessed with MODEMS Met Expectations domain.

2.3. Statistical analysis

Bivariate analysis was performed to determine associations between patient variables and surgical satisfaction. Categorical variables were assessed with Wilcoxon rank-sum test if two-groups and Kruskal-Wallis test if more than two groups. Post-hoc analysis for statistically significant categorical variables with more than two groups was performed using the Bonferroni correction. Spearman's correlation coefficient was used to assess relationships of continuous variables with SSQ-8 scores. Non-parametric testing was justified following goodness of fit testing demonstrating a minority of variables with normal distribution. Statistical tests employed were two-sided and the benchmark for statistical significance was p ​< ​0.05. Independent predictors of patient postoperative satisfaction were identified with a backwards stepwise multivariable linear regression model. Inclusion in the multivariable analysis was defined as preoperative variables significant at a level of p ​< ​0.05 in the initial bivariate analysis. JMP Pro Version 13 software was used for statistical analysis (JMP®, Version 13. SAS Institute., Cary, NC).

3. Results

Of the 361 hand and wrist surgery patients enrolled at baseline, 253 (70.1%) completed the two-year postoperative surveys. Patients had an average age of 50.4 (±16.2 years) with 138 (54.5%) being female. Mean satisfaction two-years postoperatively as assessed by SSQ-8 was 82.6 out of 100. Ceiling effect for satisfaction was observed in 58 (22.9%) patients of the cohort.

Bivariate analysis between SSQ-8 and demographic and medical history factors are summarized in Table 1 and Table 2 for categorical and continuous variables, respectively. Worse surgical satisfaction was associated with smoking status (p ​= ​0.014), preoperative opioid use (p ​= ​0.012), legal claim concerning operative indication (p ​< ​0.001), related workers compensation claim (p ​= ​0.023), and a history of clinical depression and/or anxiety (p ​< ​0.001) (Table 1). Post-hoc analysis confirmed that nonsmokers had significantly better satisfaction scores than both former and current smokers, with no significant difference between these two groups. A greater Charlson Comorbidity Index (CCI) score correlated with worse patient satisfaction (p ​= ​0.047) (Table 2). Patients with higher preoperative expectations had greater satisfaction (p ​= ​0.002). There was no significant association between satisfaction and age, number of prior orthopaedic or general surgeries, BMI, gender, race, ethnicity, or education.

Table 1.

Association of categorical patient factors and mean surgical satisfaction score (SSQ-8).

Variable	n	Mean SSQ	SD	P value
Gender
Female	138	82.1	20.3	0.89
Male	115	83.3	20.4
Race
Asian	11	89.6	10.8	0.61
Black	87	81.7	21.0
Other	9	75.5	18.3
White	141	83.0	20.7
Ethnicity
Hispanic/Latino	14	81.9	17.2	0.64
Not Hispanic/Latino	239	82.7	20.5
Education
Less than high school	14	76.1	29.4	0.21
Completed high school	125	80.3	21.2
College or more	106	85.6	18.1
Marital Status
Single	125	80.4	19.8	0.08
Married/Domestic Partnership	120	84.2	21.0
Living with possible caretaker
No	38	82.3	18.3	0.71
Yes	206	82.5	20.9
Employment
Student	25	85.0	15.5	0.77
Not employed	62	78.9	24.7
Currently Employed	159	83.4	19.4
Insurance Status
Uninsured	1	68.8	–	0.56
Government	69	82.1	20.8
Private	181	82.9	20.4
Income
<$70,000	118	80.9	20.7	0.53
>$70,000	99	83.1	21.1
Alcohol Use (monthly)
Never	76	82.9	21.4	0.78
Four or less drinks	121	82.5	20.5
More than four drinks	48	81.6	19.5
Smoking
Never Smoker	147	86.1	17.0	0.014∗
Former Smoker	60	77.8	24.8
Current smoker	39	74.8	22.2
Recreational Drugs
Do not use	233	83.2	20.1	0.10
Currently use	11	73.6	21.9
Preoperative Narcotic Use
No	168	85.0	18.5	0.012∗
Yes	82	77.2	23.7
Prior Surgery at site
No	204	83.6	18.7	0.22
Yes	46	77.3	26.0
Injury led to surgery
No	114	81.7	23.4	0.47
Yes	124	83.3	17.6
Legal Claim associated with injury
No	226	84.3	19.1	<0.001∗
Yes	16	63.7	25.2
Worker's Compensation Claim associated with injury
No	235	83.3	20.0	0.023∗
Yes	7	64.0	21.3
Depression/Anxiety
No	216	84.3	20.0	<0.001∗
Yes	37	72.5	19.4
ASA Score
I	67	85.5	14.7	0.61
II	140	79.9	22.8
III	27	82.4	19.9
IV	0	–	–

Table 2.

Association between continuous patient demographic factors and satisfaction scores.

Variable	Mean	SD	Spearman's r	P value
Age (years)	50.4	16.2	−0.01	0.85
BMI	30.2	7.9	−0.02	0.82
CCI	1.7	1.8	−0.14	0.047∗
No. Of prior ortho procedures	1.5	2.0	−0.10	0.16
No. Of lifetime surgeries	3.7	4.8	−0.11	0.13
MODEMS Preoperative Expectations	85.5	18.3	0.22	0.002∗

The most common procedure performed was carpal tunnel release (n ​= ​47, 18.6% of patients), but satisfaction did not significantly differ among the top five most frequently recorded hand surgeries by CPT code (p ​= ​0.27) (Table 3).

Table 3.

Mean surgical satisfaction score (SSQ-8) of most common procedures (CPT codes).

CPT Code (description)	n	Mean SSQ8	SD
64721	47	80.4	22.0
26055	24	80.9	24.4
25609	11	92.6	9.2
25607	9	91.7	10.2
25447	8	94.5	5.2

Mean Surgical Satisfaction Questionnaire score (SSQ-8) was analyzed by specific CPT code using Kruskal-Wallis. There was no significant difference in satisfaction among CPT codes with p ​= ​0.27.

Key: 64721- Neuroplasty and/or transposition; median nerve at carpal tunnel. 26055- Tendon sheath incision (eg, for trigger finger). 25609- Open treatment of distal radial intra-articular fracture or epiphyseal separation; with internal fixation of 3 or more fragments. 25607- Open treatment of distal radial extra-articular fracture or epiphyseal separation, with internal fixation. 25447- Arthroplasty, interposition, intercarpal or carpometacarpal joints.

Preoperative PRO measures are summarized in Table 4. Better scores on all six of the assessed PROMIS domains (PF, PI, SS, Fatigue, Anxiety, Depression) correlated with better two-year satisfaction scores (p ​< ​0.024 for all). Higher overall preoperative BMHQ score demonstrated the strongest association with better patient satisfaction among the preoperative PROs (r ​= ​0.27, p < 0.001). Baseline activity and function as measured by Tegner and Marx as well as pain via NPS at operative site were not associated with satisfaction.

Table 4.

Association between continuous preoperative patient reported factors and mean surgical satisfaction scores (SSQ-8).

Variable	Mean	SD	Spearman's r	P value
Baseline PROMIS Physical Function	46.1	9.4	0.18	0.014∗
Baseline PROMIS Pain Interference	59.1	7.4	−0.17	0.017∗
Baseline PROMIS Fatigue	50.7	9.6	−0.20	0.005∗
Baseline PROMIS Social Satisfaction	44.5	9.8	0.16	0.022∗
Baseline PROMIS Anxiety	52.9	9.0	−0.18	0.013∗
Baseline PROMIS Depression	47.9	9.1	−0.16	0.023∗
Baseline Tegner Activity	2.6	2.1	0.08	0.31
Baseline Numeric pain scale at operation site	4.7	2.9	−0.08	0.25
Baseline Marx Upper Score	48.8	28.8	0.06	0.47
Baseline BMHQ	49.4	19.9	0.27	<0.001∗

All postoperative PROs assessed were associated with two-year satisfaction and had stronger correlations compared to baseline (Table 5). Better postoperative PROMIS scores on all six domains were each associated with better satisfaction (p ​< ​0.001). Better postoperative activity measured via Tegner (p ​= ​0.002) and Marx (p ​= ​0.003) were associated with greater satisfaction. Better postoperative hand function via BMHQ was associated with greater patient satisfaction (p ​< ​0.001). Patient's expectations being met to a higher degree (MODEMS) demonstrated the strongest association among the postoperative PROs (r ​= ​0.68, p ​< ​0.001).

Table 5.

Association between continuous postoperative patient reported factors and mean surgical satisfaction scores (SSQ-8).

Variable	Mean	SD	Spearman's r	P value
2-year PROMIS Physical Function	52.6	11.0	0.38	<0.001∗
2-year PROMIS Pain Interference	50.7	10.0	−0.32	<0.001∗
2-year PROMIS Fatigue	47.1	11.2	−0.32	<0.001∗
2-year PROMIS Social Satisfaction	52.5	12.2	0.37	<0.001∗
2-year PROMIS Anxiety	50.3	11.5	−0.31	<0.001∗
2-year PROMIS Depression	47.1	10.3	−0.35	<0.001∗
2-year Tegner Activity	3.8	2.8	0.23	0.002∗
2-year Numeric pain scale at operation site	1.9	2.4	−0.52	<0.001∗
2-year Marx Upper Score	47.3	28.5	0.22	0.003∗
2-year BMHQ	79.7	20.7	0.55	<0.001∗
2-year MODEMS Met Expectations	77.3	27.1	0.68	<0.001∗

Independent preoperative predictors of patient satisfaction via multivariable linear regression are summarized in Table 6. Having no current legal claim concerning the operative indication was predictive of an estimated satisfaction score of 7.6 points higher compared to having an active legal claim (p ​= ​0.008). Never smoking was independently predictive of a better satisfaction score estimate by 4.6 points (p ​= ​0.005) compared to current or former smokers. Every 1-point increase in preoperative BMHQ predicted a 0.24-point increase in satisfaction score estimate (p ​= ​0.004). This model accounted for 15% of the variance (R² ​= ​0.15).

Table 6.

Multivariable linear regression of 2-year surgical satisfaction score (SSQ-8).

Variable	Estimate	Standard Error	P value
Legal claim (No)	7.6	2.9	0.008∗
Smoking (Never)	4.6	1.6	0.005∗
BMHQ (preoperative)	0.24	0.08	0.004∗

Adjusted R² ​= ​0.15. Using cutoff from bivariate analysis of p ​< ​0.05, the following preoperative variables were included in multivariable regression: Smoking, Preop narcotic use, Depression/Anxiety, Legal claim, Worker's compensation claim, CCI, PROMIS Preop (Physical function, pain interference, fatigue, social satisfaction, anxiety, depression), BMHQ, Preop expectations.

4. Discussion

Given the decrease in quality of life precipitated by hand and wrist pathologies, understanding drivers of patient satisfaction with hand and wrist surgery is an important consideration in assessing outcomes.^42,43 Although hand and wrist surgery satisfaction has been found to partially overlap with other surgical subspecialities, there is a need for ascertaining preoperative factors impacting satisfaction specific to the hand and wrist. To our knowledge, few studies in the hand and wrist surgery population have compared any PROMIS domains to satisfaction.⁴⁴ Here, we examined associations between satisfaction and multiple novel domains and patient specific factors in a broad hand and wrist surgery population.

The findings partially confirmed our hypothesis that better satisfaction would be associated with greater preoperative expectations, absence of depression and anxiety, and better general health (i.e. lower CCI), although these were not independently predictive on multivariable analysis. The educational status of the patient was not associated with satisfaction as hypothesized. Various other demographics, medical history factors, and PROs showed some association with satisfaction. When controlling for potential confounding variables, having no accompanying legal claim, never smoking status, and better preoperative baseline function as measured by BMHQ were all independent predictors of better patient satisfaction with hand and wrist surgery.

There is extensive literature showing that workers' compensation negatively impacts outcomes after hand and wrist procedures and surgery broadly.45, 46, 47, 48 Yet, little is known regarding any general legal claim's association with postoperative satisfaction. Katz et al. found that “involvement of an attorney” with a workers' compensation claim was associated with worse satisfaction after carpal tunnel surgery.⁴⁹ A separate study examining legal claim as a potential variable for surgical satisfaction in elective knee surgery patients found no associations between legal claims and postoperative satisfaction (as measured by both SSQ8 and Numeric Satisfaction).⁵⁰ This suggests that although having a legal claim may not negatively affect outcomes in knee surgery, involvement of the hand and wrist may increase the significance of a legal claim compared to other areas of the body. A pending legal claim may bias the patient to report worse satisfaction given the financial or other personal gain implications, especially considering the strain of a hand or wrist injury throughout everyday life. The likelihood of worse satisfaction should be kept in mind when discussing surgery for a patient with a legal claim related to their injury.

The effect of smoking on functional outcomes following hand and wrist surgery and other surgical procedures is well-documented, and this present study's finding in relation to satisfaction contributes to this understanding. Smoking has been shown to be a predictor of worse function and more symptoms postoperatively as well as associated with surgical site infections and increased revision surgeries in various surgical populations.^49,51, 52, 53 The association between smoking history and lower satisfaction is likely multidimensional, including a slowed healing process and overall worse health that may dampen a patient's outlook on the success of their intervention. However, within hand and wrist surgery, smoking status has been found to have no impact on satisfaction for carpal tunnel release and Dupuytren's contracture when studied individually.^20,23 Although carpal tunnel release was the most common procedure (47 patients), this only accounts for 18.6% of all the patients in this cohort. This suggests that even if there is no association between smoking status and postoperative satisfaction in carpal tunnel patients, this may not extend to the larger general hand and wrist surgery patient population.

To the best of our knowledge, this is the first study showing that better baseline BMHQ predicts SSQ8 scores. It is also noteworthy that all postoperative PROs were associated with surgical satisfaction, which further confirms the validity of SSQ-8. Although this relationship between functional outcomes and satisfaction has been demonstrated here and in other studies, there are still objections to utilizing patient reported satisfaction as a measure of quality of care. Arguments against the use of patient satisfaction include it being more indicative of a patient's mood regardless of the technical success of a surgery, is biased by their expectations of care without formal medical training, or that it exemplifies the interpersonal relationships that were fostered more so than functional outcomes.⁵⁴ Our findings help to refute this as SSQ-8 correlated with all two-year functional outcomes.

There are several limitations in this study. Although many patient variables were analyzed, we anticipate that others exist which influence satisfaction. Only preoperative variables were included in the multivariable analysis for predictors of satisfaction, and these only accounted for 15% of the variance as demonstrated by the R², suggesting the relevance of other patient factors not examined in this study. Retrospective analysis of prospectively acquired data has inherent design drawbacks, including that correlation does not equal causation. This study had no significant difference in satisfaction between types of hand procedures included, but these results may not be generalizable to traumatic interventions or those performed in a different population. Finally, the two-year follow up response rate was adequate (>70%), but there may be a response bias affecting our findings.

5. Conclusion

At two years following hand and wrist surgery, better patient satisfaction was independently predicted by never smoking, having no related legal claim, and better baseline Brief Michigan Hand Questionnaire scores. Future research should attempt to further characterize predictors of satisfaction following hand and wrist surgery, both at various time points and in diverse patient populations.

Funding

This work was supported by a grant from The James Lawrence Kernan Hospital Endowment Fund, Incorporated. BL1941007WS.

Declaration of competing interest

None declared