OUP accepted manuscript

Digital Tracking of Patients Undergoing Radical Cystectomy for Bladder Cancer: Daily Step Counts Before and After Surgery Within the iROC Randomised Controlled Trial

BACKGROUND

Efforts to improve recovery after radical cystectomy (RC) are needed.

OBJECTIVE

To investigate wrist-worn wearable activity trackers in RC participants.

DESIGN, SETTING, AND PARTICIPANTS

An observational cohort study was conducted within the iROC randomised trial.

INTERVENTION

Patients undergoing RC at nine cancer centres wore wrist-based trackers for 7 days (d) at intervals before and after surgery.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Step counts were compared with participant and operative features, and recovery outcomes.

RESULTS AND LIMITATIONS

Of 308 participants, 284 (92.2%) returned digital activity data at baseline (median 17 d [interquartile range: 8-32] before RC), and postoperatively (5 [5-6] d) and at weeks 5 (43 [38-43] d), 12 (94 [87-106] d), and 26 (192 [181-205] d) after RC. Compliance was affected by the time from surgery and a coronavirus disease 2019 pandemic lockdown (return rates fell to 0-7%, chi-square p < 0.001). Step counts dropped after surgery (mean of 28% of baseline), before recovering at 5 weeks (wk) (71% of baseline) and 12 wk (95% of baseline; all analysis of variance [ANOVA] p < 0.001). Baseline step counts were not associated with postoperative recovery or death. Patients with extended hospital stays had reduced postoperative step counts, with a difference of 2.2 d (95% confidence interval: 0.856-3.482 d) between the lowest third and highest two-third tertiles (linear regression analysis; p < 0.001). Additionally, they spent less time out of the hospital within 90 d of RC (80.3 vs 74.3 d, p = 0.013). Lower step counts at 5, 12, and 26 wk were seen in those seeking medical help and needing readmission (ANOVA p ≤ 0.002).

CONCLUSIONS

Baseline step counts were not associated with recovery. Lower postoperative step counts were associated with longer length of stay at the hospital and postdischarge readmissions. Studies are required to determine whether low step counts can identify patients at a risk of developing complications.

PATIENT SUMMARY

Postoperative step counts appear to be a promising tool to identify patients in the community needing medical help or readmission. More work is needed to understand which measures are most useful and how best to collect these.

Applications of artificial intelligence in surgery: clinical, technical, and governance considerations

Artificial intelligence (AI) will power many of the tools in the armamentarium of digital surgeons. AI methods and surgical proof-of-concept flourish, but we have yet to witness clinical translation and value. Here we exemplify the potential of AI in the care pathway of colorectal cancer patients and discuss clinical, technical, and governance considerations of major importance for the safe translation of surgical AI for the benefit of our patients and practices.

Artificial intelligence in surgery

Artificial intelligence (AI) is rapidly emerging in healthcare, yet applications in surgery remain relatively nascent. Here we review the integration of AI in the field of surgery, centering our discussion on multifaceted improvements in surgical care in the preoperative, intraoperative and postoperative space. The emergence of foundation model architectures, wearable technologies and improving surgical data infrastructures is enabling rapid advances in AI interventions and utility. We discuss how maturing AI methods hold the potential to improve patient outcomes, facilitate surgical education and optimize surgical care. We review the current applications of deep learning approaches and outline a vision for future advances through multimodal foundation models.

Outpatient surgery postoperative ambulation and emergency department utilization

BACKGROUND

The ability to ambulate is an important indicator for wellness and quality of life. A major health event, such as a surgery, can derail this ability, and return to preoperative walking ability is a marker for recovery. Self-reported walking measurements by patients are subject to bias, thus wearable technology such as activity monitors have risen in popularity. We evaluated postoperative ambulation using an accelerometer in outpatient general surgery procedures with the hypothesis that those patients with less postoperative ambulation were at risk for adverse outcomes.

METHODS

A retrospective review of patients undergoing outpatient abdominal surgeries from November 2016 to July 2019 at a Veteran Affairs Medical Center. Patients wore an accelerometer preoperatively and postoperatively to measure their ambulation (steps/day). Outcome measures were 30-day readmissions and Emergency Department (ED) utilization. Postoperative ambulation was defined as daily percentages of their preoperative baseline. Patients without preoperative baseline data, > 3 missing days or any missing days prior to reaching baseline were excluded.

RESULTS

One-hundred-six patients underwent outpatient abdominal surgery. Twenty-two patients were excluded. Patients stratified into adult (18-64 years, 44 patients, 52%) and geriatric (≥ 65 years, 40 patients, 48%) cohorts. Geriatric patients were less likely to meet their preoperative baseline by postoperative day 7, 35% vs 61%, p = 0.016. Adult patients who failed to meet their preoperative baseline in first postoperative week had higher ED utilization; 4 (24%) vs 1 (4%), p = 0.04. Geriatric patients who failed to meet their baseline trended toward increased ED utilization; 5 (19%) vs. 1 (7%), p = 0.31.

CONCLUSION

Patients aged < 65 who fail to return to their preoperative daily step count within one week of outpatient abdominal surgery are 6× more likely to be seen in the ED. Postoperative ambulation may be able to predict ED utilization and recovery after outpatient surgery.

Wearable Sensors as a Preoperative Assessment Tool: A Review

Surgery is a common first-line treatment for many types of disease, including cancer. Mortality rates after general elective surgery have seen significant decreases whilst postoperative complications remain a frequent occurrence. Preoperative assessment tools are used to support patient risk stratification but do not always provide a precise and accessible assessment. Wearable sensors (WS) provide an accessible alternative that offers continuous monitoring in a non-clinical setting. They have shown consistent uptake across the perioperative period but there has been no review of WS as a preoperative assessment tool. This paper reviews the developments in WS research that have application to the preoperative period. Accelerometers were consistently employed as sensors in research and were frequently combined with photoplethysmography or electrocardiography sensors. Pre-processing methods were discussed and missing data was a common theme; this was dealt with in several ways, commonly by employing an extraction threshold or using imputation techniques. Research rarely processed raw data; commercial devices that employ internal proprietary algorithms with pre-calculated heart rate and step count were most commonly employed limiting further feature extraction. A range of machine learning models were used to predict outcomes including support vector machines, random forests and regression models. No individual model clearly outperformed others. Deep learning proved successful for predicting exercise testing outcomes but only within large sample-size studies. This review outlines the challenges of WS and provides recommendations for future research to develop WS as a viable preoperative assessment tool.

Surveillance of high-risk early postsurgical patients for real-time detection of complications using wireless monitoring (SHEPHERD study): results of a randomized multicenter stepped wedge cluster trial

Background

Vital signs measurements on the ward are performed intermittently. This could lead to failure to rapidly detect patients with deteriorating vital signs and worsens long-term outcome. The aim of this study was to test the hypothesis that continuous wireless monitoring of vital signs on the postsurgical ward improves patient outcome.

Methods

In this prospective, multicenter, stepped-wedge cluster randomized study, patients in the control group received standard monitoring. The intervention group received continuous wireless monitoring of heart rate, respiratory rate and temperature on top of standard care. Automated alerts indicating vital signs deviation from baseline were sent to ward nurses, triggering the calculation of a full early warning score followed. The primary outcome was the occurrence of new disability three months after surgery.

Results

The study was terminated early (at 57% inclusion) due to COVID-19 restrictions. Therefore, only descriptive statistics are presented. A total of 747 patients were enrolled in this study and eligible for statistical analyses, 517 patients in the control group and 230 patients in the intervention group, the latter only from one hospital. New disability at three months after surgery occurred in 43.7% in the control group and in 39.1% in the intervention group (absolute difference 4.6%).

Conclusion

This is the largest randomized controlled trial investigating continuous wireless monitoring in postoperative patients. While patients in the intervention group seemed to experience less (new) disability than patients in the control group, results remain inconclusive with regard to postoperative patient outcome due to premature study termination.

Clinical trial registration

ClinicalTrials.gov, ID: NCT02957825.

Hypotension duration and vasopressor requirements following intrathecal oxytocin for Total hip arthroplasty: Secondary analysis of a randomized controlled trial

INTRODUCTION

A recent publication investigating intrathecal oxytocin, 100 μg, administered immediately prior to a spinal anesthetic in patients undergoing primary total hip arthroplasty surgery demonstrated a reduction in disability for 3-weeks, increased walking distance at 8-weeks, and earlier opioid cessation. This secondary analysis study was undertaken to assess the acute cardiovascular safety and analgesic efficacy of intrathecal oxytocin in this study population.

METHODS

90 patients were included in the analysis (44 randomized to spinal oxytocin and 46 to placebo [saline]). Data collected prospectively during the previously published study were supplemented with additional retrospectively collected data. The primary outcomes were comparisons of the duration of hypotension (minutes with mean arterial pressure < 65 mmHg) and cumulative vasopressor requirements during the initial 60 min following spinal placement. Secondary outcomes included hypotension durations and vasopressor requirements at later time points, perioperative fluid administration, physical therapy metrics, time to first opioid administration, cumulative opioid consumption through 24 h, and verbal pain scores through 24 h.

RESULTS

The duration of hypotension during the first 60 min following spinal placement did not differ between intrathecal oxytocin and placebo groups (12.2 ± 10.7 vs 14.0 ± 13.0 min, respectively; p = 0.476). There was also no difference in cumulative vasopressor requirements (1303 ± 883 vs 1156 ± 818 μg [phenylephrine equivalents]; p = 0.413) during that time period. No group differences were found for any of the investigated secondary outcomes.

CONCLUSION

The administration of 100 μg of intrathecal oxytocin does not significantly impact the duration of hypotension or the need for vasopressor agents when given as a component of a spinal anesthetic. The oxytocin and placebo groups also did not differ in regards to physical therapy related metrics, time to first opioid administration, cumulative opioids at 24-h, or pain scores through 24-h. What is already known on this topic: Rapid intravenous oxytocin causes hypotension after cesarean delivery, but intrathecal oxytocin does not cause hypotension in healthy volunteers.

WHAT THIS STUDY ADDS

Compared to saline control, intrathecal oxytocin, 100 μg did not increase the duration of hypotension or vasopressor requirements in patients during total hip arthroplasty. How this study might affect research, practice, or policy: Lack of hypotension from intrathecal oxytocin in this study supports future investigations to further explore its potential benefits, in terms of both analgesia and functional recovery following surgery.

Comparison of postoperative mobilization measurements by activPAL versus Johns Hopkins Highest Level of Mobility scale after major abdominal surgery

BACKGROUND

The Johns Hopkins Highest Level of Mobility scale is a validated tool for assessing patient mobility in the hospital. It has excellent inter-rater and test-retest reliabilities, but it is unknown how accurately Johns Hopkins Highest Level of Mobility documentation reflects the patients' mobility performance in the immediate postoperative period compared to objective measures such as accelerometers.

METHODS

In this single-center observational study, consented adults undergoing open abdominal surgery wore a research-grade accelerometer, activPAL, starting immediately postoperatively until hospital discharge or up to 7 days. We collected the Johns Hopkins Highest Level of Mobility scores documented by hospital staff via retrospective chart review and evaluated their accuracy in describing the type, frequency, and volume of postoperative out-of-bed mobilization using the activPAL as the criterion.

RESULTS

We analyzed data from 56 participants. The activPAL showed that participants spent 97.7% of their time lying in bed or sitting in a chair. Meanwhile, the Johns Hopkins Highest Level of Mobility documentation of preambulatory activities (scores 1-5) was rare. The activPAL detected 4 times more out-of-bed mobilization than routine Johns Hopkins Highest Level of Mobility documentation. Whereas the frequency of activPAL-measured out-of-bed mobilization increased steadily to a median of 9 sessions by postoperative day 6, the number of Johns Hopkins Highest Level of Mobility documentation remained around twice daily. ActivPAL measurements demonstrated that Johns Hopkins Highest Level of Mobility documentation of ambulatory sessions (scores 6-8) was accurate.

CONCLUSIONS

We found that routine Johns Hopkins Highest Level of Mobility documentation did not accurately detect preambulatory activities or the overall frequency of out-of-bed mobility sessions, poorly reflecting the highly sedentary behaviors of the acute postoperative inpatients and highlighting the need to improve clinical documentation or use alternative methods to track postoperative mobilization.

Implementing Wearable Sensors for Clinical Application at a Surgical Ward: Points to Consider before Starting

Incorporating technology into healthcare processes is necessary to ensure the availability of high-quality care in the future. Wearable sensors are an example of such technology that could decrease workload, enable early detection of patient deterioration, and support clinical decision making by healthcare professionals. These sensors unlock continuous monitoring of vital signs, such as heart rate, respiration rate, blood oxygen saturation, temperature, and physical activity. However, broad and successful application of wearable sensors on the surgical ward is currently lacking. This may be related to the complexity, especially when it comes to replacing manual measurements by healthcare professionals. This report provides practical guidance to support peers before starting with the clinical application of wearable sensors in the surgical ward. For this purpose, the Non-Adoption, Abandonment, Scale-up, Spread, and Sustainability (NASSS) framework of technology adoption and innovations in healthcare organizations is used, combining existing literature and our own experience in this field over the past years. Specifically, the relevant topics are discussed per domain, and key lessons are subsequently summarized.

Impact of Digital Health upon the Surgical Patient Experience: The Patient as Consumer

The adoption of digital health services in surgical care delivery is changing the patient experience. The goal of patient-generated health data monitoring incorporated with patient-centered education and feedback is to optimally prepare patients for surgery and personalize postoperative care to improve outcomes that matter to both patients and surgeons. Challenges include the need for the adoption of new methods for implementation and evaluation and equitable application of surgical digital health interventions, with considerations for accessibility as well as the development of new diagnostics and decision support that include the needs and characteristics of all populations served.

Risk for Poor Post-Operative Quality of Life Among Wearable Use Subgroups in an All of Us Research Cohort

The objective of this research was to build and assess the performance of a prediction model for post-operative recovery status measured by quality of life among individuals experiencing a variety of surgery types. In addition, we assessed the performance of the model for two subgroups (high and moderately consistent wearable device users). Study variables were derived from the electronic health records, questionnaires, and wearable devices of a cohort of individuals with one of 8 surgery types and that were part of the NIH All of Us research program. Through multivariable analysis, high frailty index (OR 1.69, 95% 1.05-7.22, p<0.006), and older age (OR 1.76, 95% 1.55-4.08, p<0.024) were found to be the driving risk factors of poor recovery post-surgery. Our logistic regression model included 15 variables, 5 of which included wearable device data. In wearable use subgroups, the model had better accuracy for high wearable users (81%). Findings demonstrate the potential for models that use wearable measures to assess frailty to inform clinicians of patients at risk for poor surgical outcomes. Our model performed with high accuracy across multiple surgery types and were robust to variable consistency in wearable use.

Telemonitoring for perioperative care of outpatient bariatric surgery: Preference-based randomized clinical trial

IMPORTANCE

Implementation of bariatric surgery on an outpatient basis is hampered by concerns about timely detection of postoperative complications. Telemonitoring could enhance detection and support transition to an outpatient recovery pathway.

OBJECTIVE

This study aimed to evaluate non-inferiority and feasibility of an outpatient recovery pathway after bariatric surgery, supported by remote monitoring compared to standard care.

DESIGN

Preference-based non-inferiority randomized trial.

SETTING

Center for obesity and metabolic surgery, Catharina hospital Eindhoven, the Netherlands.

PARTICIPANTS

Adult patients scheduled for primary gastric bypass or sleeve gastrectomy.

INTERVENTIONS

Same-day discharge with one week ongoing Remote Monitoring (RM) of vital parameters or Standard Care (SC) with discharge on postoperative day one.

MAIN OUTCOMES

Primary outcome was a thirty-day composite Textbook Outcome score encompassing mortality, mild and severe complications, readmission and prolonged length-of-stay. Non-inferiority of same-day discharge and remote monitoring was accepted below the selected margin of 7% upper limit of confidence interval. Secondary outcomes included admission duration, post-discharge opioid use and patients' satisfaction.

RESULTS

Textbook Outcome was achieved in 94% (n = 102) in RM versus 98% (n = 100) in SC (RR 2.9; 95% CI, 0.60-14.23, p = 0.22). The non-inferiority margin was exceeded which is a statistically inconclusive result. Both Textbook Outcome measures were above Dutch average (5% RM and 9% SC). Same-day discharge reduced hospitalization days by 61% (p<0.001) and by 58% with re-admission days included (p<0.001). Post-discharge opioid use and satisfaction scores were equal (p = 0.82 and p = 0.86).

CONCLUSION

In conclusion, outpatient bariatric surgery supported with telemonitoring is clinically comparable to standard overnight bariatrics in terms of textbook-outcome. Both approaches reached primary endpoint results above Dutch average. However, statistically the outpatient surgery protocol was neither inferior, nor non-inferior to the standard pathway. Additionally, offering same-day discharge reduces the total hospitalization days while maintaining patient satisfaction and safety.

Characterising nationwide reasons for unplanned hospital readmission after colorectal cancer surgery

BACKGROUND

Readmissions after colorectal cancer surgery are common, despite advancements in surgical care, and have a significant impact on both individual patients and overall healthcare costs. The aim of this study was to determine the 30-and 90 days readmission rate after colorectal cancer surgery, and to investigate the risk factors and clinical reasons for unplanned readmissions.

METHOD

A multicenter, population-based study including all patients discharged after index colorectal cancer resection from 2010 to 2020 in Aotearoa New Zealand (AoNZ) was completed. The Ministry of Health National Minimum Dataset was used. Rates of readmission at 30 days and 90 days were calculated. Mixed-effect logistic regression models were built to investigate factors associated with unplanned readmission. Reasons for readmission were described.

RESULTS

Data were obtained on 16,885 patients. Unplanned 30-day and 90-day hospital readmission rates were 15.1% and 23.7% respectively. The main readmission risk factors were comorbidities, advanced disease, and postoperative complications. Hospital level variation was not present. Despite risk adjustment, R² value of models was low (30 days: 4.3%, 90 days: 5.2%). The most common reasons for readmission were gastrointestinal causes (32.1%) and wound complications (14.4%). Rates of readmission did not improve over the 11 years study period (p = 0.876).

CONCLUSION

Readmissions following colorectal resections in AoNZ are higher than other comparable healthcare systems and rates have remained constant over time. While patient comorbidities and postoperative complications are associated with readmission, the explanatory value of these variables is poor. To reduce unplanned readmissions, efforts should be focused on prevention and early detection of post-discharge complications.

Passively Captured Interpersonal Social Interactions and Motion From Smartphones for Predicting Decompensation in Heart Failure: Observational Cohort Study

Background

Heart failure (HF) is a major cause of frequent hospitalization and death. Early detection of HF symptoms using smartphone-based monitoring may reduce adverse events in a low-cost, scalable way.

Objective

We examined the relationship of HF decompensation events with smartphone-based features derived from passively and actively acquired data.

Methods

This was a prospective cohort study in which we monitored HF participants’ social and movement activities using a smartphone app and followed them for clinical events via phone and chart review and classified the encounters as compensated or decompensated by reviewing the provider notes in detail. We extracted motion, location, and social interaction passive features and self-reported quality of life weekly (active) with the short Kansas City Cardiomyopathy Questionnaire (KCCQ-12) survey. We developed and validated an algorithm for classifying decompensated versus compensated clinical encounters (hospitalizations or clinic visits). We evaluated models based on single modality as well as early and late fusion approaches combining patient-reported outcomes and passive smartphone data. We used Shapley additive explanation values to quantify the contribution and impact of each feature to the model.

Results

We evaluated 28 participants with a mean age of 67 years (SD 8), among whom 11% (3/28) were female and 46% (13/28) were Black. We identified 62 compensated and 48 decompensated clinical events from 24 and 22 participants, respectively. The highest area under the precision-recall curve (AUCPr) for classifying decompensation was with a late fusion approach combining KCCQ-12, motion, and social contact features using leave-one-subject-out cross-validation for a 2-day prediction window. It had an AUCPr of 0.80, with an area under the receiver operator curve (AUC) of 0.83, a positive predictive value (PPV) of 0.73, a sensitivity of 0.77, and a specificity of 0.88 for a 2-day prediction window. Similarly, the 4-day window model had an AUC of 0.82, an AUCPr of 0.69, a PPV of 0.62, a sensitivity of 0.68, and a specificity of 0.87. Passive social data provided some of the most informative features, with fewer calls of longer duration associating with a higher probability of future HF decompensation.

Conclusions

Smartphone-based data that includes both passive monitoring and actively collected surveys may provide important behavioral and functional health information on HF status in advance of clinical visits. This proof-of-concept study, although small, offers important insight into the social and behavioral determinants of health and the feasibility of using smartphone-based monitoring in this population. Our strong results are comparable to those of more active and expensive monitoring approaches, and underscore the need for larger studies to understand the clinical significance of this monitoring method.

Outcomes of Vital Sign Monitoring of an Acute Surgical Cohort With Wearable Sensors and Digital Alerting Systems: A Pragmatically Designed Cohort Study and Propensity-Matched Analysis

Background: The implementation and efficacy of wearable sensors and alerting systems in acute secondary care have been poorly described.

Objectives: to pragmatically test one such system and its influence on clinical outcomes in an acute surgical cohort.

Methods: In this pragmatically designed, pre-post implementation trial, participants admitted to the acute surgical unit at our institution were recruited. In the pre-implementation phase (September 2017 to May 2019), the SensiumVitals™ monitoring system, which continuously measures temperature, heart, and respiratory rates, was used for monitoring alongside usual care (intermittent monitoring in accordance with the National Early Warning Score 2 [NEWS 2] protocol) without alerts being generated. In the post-implementation phase (May 2019 to March 2020), alerts were generated when pre-established thresholds for vital parameters were breached, requiring acknowledgement from healthcare staff on provided mobile devices. Hospital length of stay, intensive care use, and 28-days mortality were measured. Balanced cohorts were created with 1:1 ‘optimal’ propensity score logistic regression models.

Results: The 1:1 matching method matched the post-implementation group (n = 141) with the same number of subjects from the pre-implementation group (n = 141). The median age of the entire cohort was 52 (range: 18–95) years and the median duration of wearing the sensor was 1.3 (interquartile range: 0.7–2.0) days. The median alert acknowledgement time was 111 (range: 1–2,146) minutes. There were no significant differences in critical care admission (planned or unplanned), hospital length of stay, or mortality.

Conclusion: This study offered insight into the implementation of digital health technologies within our institution. Further work is required for optimisation of digital workflows, particularly given their more favourable acceptability in the post pandemic era. Clinical trials registration information: ClinicalTrials.gov Identifier: NCT04638738.