Prospective cohort study investigating quality of life outcomes following multi-speciality robotic-assisted surgery.

Robotic-assisted surgery (RAS) has been increasingly incorporated into surgical practice over the past two decades with increasingly complex procedures performed with a minimally invasive approach.[ 1 ] Emerging evidence suggests that RAS has clinical advantages over open or laparoscopic surgery, with respect to greater precision, lower blood loss and transfusion rates, reduced pain, shorter length of stay (LOS), lower incidence of complications and more rapid post-operative recovery.[ 2 3 4 5 ] However, RAS is associated with longer operative times, not universally available owing to capital and consumable expenditure, and requires a steep learning curve for complex robotic procedures.[ 6 7 ] Given the significant operating costs associated with the current robotic systems,[ 1 8 ] the potential impact of RAS on both clinical and patient-reported outcomes needs to be evaluated. Current evidence on quality of life (QoL) outcomes in patients undergoing RAS is limited and mixed.[ 7 9 ] Some studies found no significant differences between RAS and conventional approaches in QoL,[ 9 10 11 12 13 ] while others showed improvement in physical, functional and general health scores.[ 5 14 15 16 ] Most studies have relied on measuring outcomes with disease-specific functional assessment questionnaires[ 5 7 10 11 ] or had small sample sizes.[ 17 18 19 ] Some studies lacked a baseline measure of QoL or did not have a long-term post-operative follow-up.[ 10 11 ] With the increasing incorporation of RAS into practice, particularly within the public sector, more reliable evidence on QoL outcomes of RAS is needed to inform patient and health-care provider decision-making regarding the efficacy of the technology.[ 20 ] A wide spectrum of QoL should be considered to provide a global assessment of changes in patient QoL that encompasses social and emotional functioning as well as physical and mental well-being after RAS. A generic measure will also allow comparisons in QoL across a range of disease conditions and treatments. Given the goal of surgical procedures is to treat patients whilst at the same time maintaining or improving their physical outcomes,[ 21 22 ] QoL should be an important measure to be considered in the evaluation of RAS. This study aims to examine the direction and magnitude of changes in QoL outcomes for a multi-speciality cohort of consecutive patients undergoing RAS in a large public tertiary referral hospital in Sydney Australia.

The study sample included patients undergoing urologic (robotic-assisted radical prostatectomy), cardiothoracic (stepwise partial robotic-assisted coronary artery bypass grafting and robotic-assisted lung resection), colorectal (robotic-assisted rectal resection) or benign gynaecological (robotic-assisted hysterectomy and endometriosis) RAS at Royal Prince Alfred Hospital (RPAH) between June 2016 and January 2020. RPAH is one of the first public tertiary referral hospitals in Australia to have the generation Da Vinci Xi Surgical System. Ethics approval was obtained from the RPAH Ethics Committee. Signed consent was obtained from all participants. QoL outcomes were evaluated using the Short-Form 36 (SF-36 version 2). The SF-36 is one of the most widely used generic measures of QoL for routine assessment of care and health outcomes in adult patients.[ 23 ] The 36 items are grouped into eight sub-domains, including physical functioning, role limitations due to physical health problems, bodily pain, general health perception, vitality, role limitations due to emotional problems, social functioning and mental health, with each domain scored on a 0–100 scale. A higher score indicates better QoL. These domains can be combined into two main components scores, physical and mental summary scores, using a norm-based approach.[ 24 ] SF-36 scores were also transformed into a utility index on a 0 (death)-1 (full health) scale using the algorithm developed from the United Kingdom population.[ 25 ] Patients were given the SF-36 questionnaires to be completed through confidential self-administration and returned before surgery (baseline), and at 6 weeks and 6 months post-surgery. Patients who did not return the questionnaires were contacted by telephone and given the option to complete the questionnaires on the phone or receive a prepaid envelope to return the questionnaires. For missing values in SF-36 items, an imputation procedure was performed within each domain based on the guidelines of the SF-36.[ 23 ] Confounding factors considered in the regression adjustment included demographics (i.e. age, gender, country of birth and caring responsibilities), socioeconomics (i.e. area socioeconomic index and employment status), illness (i.e. cancer, urgent admission, intensive care unit stay and LOS) and health-care care utilisation since being discharged from hospital (i.e. treatment, visit to the emergency department [ED], visit to the general practitioner, surgeon, specialist, senior hospital nurse, community nurse, or an allied health professional and paid/unpaid home care). Information was extracted from a comprehensive surgical robotics programme database (BEST). Means (standard deviation) and frequencies (percentage) of patient characteristics by surgical specialities were reported for continuous and categorical factors, respectively. Changes in the direction, magnitude and statistical significance of QoL outcomes (i.e. physical and mental summary scores, utility index and sub-domains of the SF-36) by surgical specialities were assessed using mixed effects linear regressions. Random effects were identified as random intercepts at the level of individual patients. Standard errors were clustered at the individual level. Estimates from models without (unadjusted) and with (adjusted) patient characteristics were reported. In the adjusted regression models, analyses were controlled for demographic, socioeconomic, illness, treatment and health-care utilisation factors. All analyses were conducted using Stata 14.2.

A total cohort of 254 patients were included, including 154 (61%) urologic (i.e. 150 [59%] prostatectomy and 4 [2%] partial nephrectomy), 36 (14%) cardiothoracic (i.e. 31 [12%] coronary artery bypass grafting and 5 [2%] thoracic resections), 24 (9%) colorectal (i.e. rectal resections) and 40 (16%) benign gynaecological RAS (i.e. hysterectomy and endometriosis resections). Of these patients, 229 (90.2%) returned the 6 weeks’ post-operative questionnaire (136 [59%] urologic, 32 [14%] cardiothoracic, 22 [10%] colorectal and 39 [17%] benign gynaecological) and 228 (89.8%) returned the 6 months questionnaire (153 [67%] urologic, 20 [9%] cardiothoracic, 16 [7%] colorectal and 39 [17%] benign gynaecological). Patient characteristics are reported in Table 1 . Patients were on average 58.8 years of age ranging from 18 to 87, 75.1% (193) were male, 51.6% (174) were employed preoperatively, and 45.7% (116) were in the highest quartile of the area socioeconomics index. Cancer surgery predominated in urology (152, 98.1%) and colorectal (19, 79.2%) RAS. Post-operative LOS was longer for colorectal (13.3 days) and cardiothoracic (11.6 days) RAS than for urologic (2.5 days) and benign gynaecological (1.7 days) RAS. Characteristics of the included sample SD: Standard deviation, ICU: Intensive care unit, ED: Emergency department, GP: General practitioner Figure 1a - c displays the descriptive trajectory of the mean and 95% confidence intervals (95% CIs) for the SF-36 physical and mental summary scores and utility index at baseline, 6 weeks and 6 months postoperatively. Physical summary scores for urologic and colorectal RAS returned to baseline levels within 6 months postoperatively, despite an initial decrease at 6 weeks’ post-operative; while physical summary scores for cardiothoracic and benign gynaecological RAS had consistent improvements following surgery. Mental summary scores for cardiothoracic and benign gynaecological RAS increased at postoperatively when compared to baseline; mental summary scores for urologic and cardiothoracic RAS returned to baseline levels at post-operative 6 months. Utility index increased at 6 months postoperatively compared to baseline scores for urologic, colorectal and benign gynaecological RAS. Trajectories for sub-domains are presented in Figure 2 . Quality of life outcomes following robotic-assisted surgery. The mean and 95% CIs are displayed for physical and mental summary scores, on 0-100 scale, and utility index, on 0-1 scale. A higher score indicates better health. (a) Physical component scores. (b) Mental component scores. (c) Utility index Quality of life sub-domain following robotic-assisted surgery The results from unadjusted mixed-effects linear regressions for urologic RAS show a decrease in physical summary score (−3.79; 95% CI: −5.28, −2.29) and no significant changes in mental summary score and utility at 6 weeks postoperatively [ Table 2 ]. However, there was a significant improvement in the mental health sub-domain at 6 months postoperatively (2.22; 95% CI: 0.03, 4.42). The results from adjusted regressions [ Table 3 ] remain largely consistent after controlling for sociodemographic, illness and service use covariates. Unadjusted mixed effects models for changes in quality of life compared to baseline * P <0.05. Multilevel mixed-effects linear regressions were performed, with clustered SE; coefficients and standard errors are reported; positive coefficients indicate increases in QoL whereas negative coefficients indicate decreases in QoL. QoL: Quality of life, SE: Standard error Adjusted mixed effects models for changes in quality of life compared to baseline by surgical speciality * P <0.05. Mixed effects linear regressions were performed, adjusting for age, gender, country of birth, caring responsibility, SEIFA quartiles, work status, cancer status, admission urgency, ICU status, LOS and health service use factors, with clustered SE; coefficients and SEs are reported; positive coefficients indicate increases in QoL whereas negative coefficients indicate decreases in QoL. QoL: Quality of life, ICU: Intensive care unit, LOS: Length of stay, SE: Standard error, SEIFA: Socioeconomic indexes for areas While there were no significant changes in physical and mental summary scores and utility for cardiothoracic RAS at 6 weeks postoperatively, significant improvements were observed for physical summary score (5.20; 95% CI: 1.03, 9.36) and bodily pain sub-domain (6.77; 95% CI: 1.31, 12.24) 6 months after surgery [ Table 2 ]. After controlling for sociodemographic, illness and health-care covariates, no significant changes in QoL were revealed [ Table 3 ]. There was a decrease in the physical summary score at 6 weeks postoperatively (-4.70; 95%CI: -8.23, −1.16), but significant increases in the mental summary score at 6 weeks (3.88; 95% CI: 0.15, 7.61) and 6 months postoperatively (6.69; 95% CI: 2.63, 10.76) [ Table 2 ]. In addition, significant improvement was found at 6 months in sub-domain role physical (3.64; 95% CI: 0.89, 6.39), vitality (4.57; 95% CI: 1.63, 7.52), social functioning (7.05; 95% CI: 2.38, 11.73), role emotional (3.29; 95% CI: 0.17, 6.41) and mental health (6.84; 95% CI: 2.01, 11.67). Similar results were found in adjusted regressions [ Table 4 ]. Adjusted mixed effects models for changes in quality of life * P <0.10, ** P <0.05, *** P <0.01. Mixed effects linear regressions were performed for all patients across surgical specialties, controlling time indicators, specialty indicators and other covariates, with clustered SE; coefficients, SE and AIC are reported. AIC: Akaike’s information criteria, SE: Standard error, ICU: Intensive care unit, ED: Emergency department, GP: General practitioner, LOS: Length of stay, SEIFA: Socioeconomic indexes for areas Improvements were statistically significant for physical summary scores at 6 weeks postoperatively (3.56; 95% CI: 0.89, 6.23); mental summary scores at 6 weeks (6.57; 95% CI: 3.34, 9.79) and 6 months postoperatively (9.94; 95CI: 6.15, 13.73) and utility index at 6 weeks (0.08; 95% CI: 0.04, 0.11) and 6 months postoperatively (0.08; 95% CI: 0.04, 0.12) [ Table 2 ]. Increases were also consistently observed across sub-domains at 6 months for role physical (5.22; 95% CI: 1.71, 8.72), bodily pain (8.74; 95% CI: 4.16, 13.31), general health (5.08; 95% CI: 1.33, 8.82), vitality (9.05; 95% CI: 5.64, 12.46), social functioning (9.16; 95% CI: 4.65, 13.68), role emotional (6.87; 95% CI: 3.30, 10.44) and mental health (10.32; 95% CI: 6.46, 14.19). Significant and large improvements were confirmed in adjusted regressions [ Table 3 ]. The mental summary score increased with age while the physical summary score decreased with age [ Table 4 ]. Residing in relatively socioeconomically advantaged areas and fewer LOS were associated with higher QoL scores. Having attended a hospital ED since being discharged, having had contacts with a general practitioner, a hospital nurse, a community nurse, or having received unpaid home help were associated with lower QoL scores, which likely reflected the need for extra healthcare associated with illness severity.

This study provides a comprehensive assessment of QoL in urologic, cardiothoracic, colorectal and benign gynaecological RAS before and after surgery using a large prospective cohort at a large public tertiary referral hospital in Sydney, Australia. The results show that there was a significant improvement in QoL following RAS, indicating potential benefits in the use of RAS. Consistent with previous literature,[ 26 ] there was variation in QoL improvements across surgical specialities. For urological RAS, there was an initial decrease in physical summary scores 6 weeks postoperatively that were likely associated with postsurgery recovery and incontinence; however, the scores improved back to baseline level at 6 months postoperatively as continence improved. For cardiothoracic RAS, significant improvements in physical health and bodily pain were observed at 6 months postoperatively, although the significance disappeared once other covariates were controlled for. For colorectal RAS, physical summary scores had an initial decrease at 6 weeks postoperatively followed by a recovery to baseline at 6 months. For benign gynaecological RAS, physical summary scores and numerous sub-domains had immediate increases at 6 weeks and consistent increases at 6 months postoperatively. For urological RAS, there was an increase in the mental health sub-domain at 6 months, which may be due to the fact that following the feeling of being stressed and anxious about the cancer diagnosis before surgery, the mental health of patients improved on completion of surgery and recovery. For cardiothoracic RAS, mental summary scores improved 6 months postoperatively, although no significant changes were observed once controlling for other covariates. For benign gynaecological RAS, mental summary scores and numerous sub-domains had persistent increases at 6 weeks and 6 months postoperatively, which could be explained by the recovery of gynaecological patients who were symptomatic at the time of presentation (e.g. endometriosis). For colorectal RAS, mental summary scores and several QoL sub-domains also had significant improvements post 6 weeks and 6 months postoperatively. The improvement for mental health after RAS could be partially attributed to patients having been cured or treated for their conditions and thus the relief and joy they felt, or health professionals engaging with patients and informing them of improved clinical outcomes expected using RAS. The utility index had consistent increases at 6 weeks and 6 months for benign gynaecological RAS, resulting in significant improvements across physical and mental domains. Existing evidence on changes in generic QoL outcomes associated with RAS is somewhat limited. Previous studies are in agreement with the findings of the current study, reporting improvements in QoL outcomes after surgery. Studies using the SF-12 survey reported a significant improvement in QoL following robotic-assisted inguinal hernia repair.[ 27 ] While another study reported return of QoL to baseline levels after 12 weeks following robotic-assisted partial nephrectomy.[ 26 ] Other studies adopting the EQ-5D survey revealed that QoL returned to the pre-operative levels at 5 weeks after robotic-assisted hysterectomy.[ 28 ] Patients undergoing robotic-assisted hysterectomy had larger improvement in QoL outcomes 6–8 weeks postoperatively when compared to conventional laparoscopic hysterectomy for benign disease.[ 29 ] Further, when compared to laparoscopic and open surgical approaches, RAS have demonstrated equivalent or superior QoL outcomes.[ 30 31 32 ] However, other recent studies also described insufficient evidence to suggest significant improvements in QoL outcomes after RAS.[ 13 33 34 ] The current study contributes to the growing research on the evaluation of RAS by describing the medium-term QoL trajectories in a large cohort of patients undergoing multi-specialty RAS. This is a strength of our study, which allowed increased precision of estimates and control of heterogeneity, when compared to the previous literature that was drawn on shorter-term outcomes (e.g. 6 weeks–3 months),[ 15 28 33 ] small samples (e.g. <25 or 50 patients),[ 17 18 19 ] and/or a single specialty cohort.[ 12 15 22 ] Further, instead of disease-specific functional measures,[ 10 11 22 35 36 37 ] the current study utilised a generic measure of QoL that allowed assessments on physical, social and psychological well-being of patients, and enabled comparisons across disease conditions or treatments.[ 23 ] The study has some limitations. First, there is a lack of randomised control trial design and a comparator surgical approach. The current prospective cohort study focused on within-group changes and lacked information on incremental QoL between RAS and other comparator groups within a controlled study design. Second, a longer follow-up period would allow assessment of long-term changes in the well-being of patients who underwent RAS, which has not been sufficiently evaluated to date. QoL should be considered one of the major long-term outcomes after surgical procedures. Future work should also incorporate costing information on RAS to enable a cost-effectiveness evaluation for different surgical specialties. This prospective cohort study suggested significant improvements in QoL following urologic, cardiothoracic, colorectal, or benign gynaecological RAS, demonstrating potential efficacy in utilising RAS technology. The degree of improvement in physical and mental domains varied with surgical specialities. Nil. There are no conflicts of interest.