KEY POINTS
Question: Can pedometers and other wearable monitoring devices be used to identify patients at risk for longer hospital stay and developing postoperative complications?
Findings: Analyzed studies showed that preoperative pedometer readings correlated with postoperative stay and complication rate, while inpatient postoperative pedometer recordings correlated with postdischarge complications.
Meaning: Perioperative pedometer data demonstrated consistent and biologically plausible association with perioperative outcomes, thus warranting further investigation as a perioperative risk stratification tool.
A pedometer is an electromechanical device that detects motion pattern and estimates the number of steps taken by the wearer over time. Advances in mobile technology have integrated pedometer devices into a wide range of devices, including smartphones, smartwatches, and other wearable activity trackers.
It is widely accepted that patients’ functional capacity has a significant influence on perioperative outcomes. Patients who require assistance with daily activities are at considerably higher risk of postoperative cardiopulmonary complications.1,2 Sedentary behavior preoperatively is similarly associated with higher risk of postoperative complications.3 Low physical activity level is a feature of frailty, which is associated with considerable perioperative morbidity and mortality.4
Subjective patient histories may be inaccurate, while bedside walking tests only offer a snapshot of a patient’s exercise tolerance.5,6 Cardiopulmonary exercise testing (CPET) could provide an objective measurement of patients’ exercise tolerance,7 but it is resource-intensive. Assessing patients’ functional recovery after surgery presents similar challenges.
The use of pedometer data for the evaluation of functional capacity offers several advantages. Patients could be monitored remotely for an extended period of time, which minimizes the influence of any acute events on the assessment and allows for evaluation of activity changes over time. Studies have also reported a moderate-to-strong correlation between daily step count and CPET results.8,9 Pedometer devices do not require direct interaction with health care professionals, something particularly important during the coronavirus pandemic. These characteristics make the pedometer a potentially useful tool for functional capacity assessment, both for perioperative risk stratification and predicting postoperative recovery. We, therefore, conducted this systematic review to summarize the clinical evidence on the perioperative use of wearable pedometer devices.
METHODS
Study Objectives
The aim of our study was to evaluate the clinical utility of pedometers during the perioperative period. We considered all relevant prospective and retrospective studies of patients who underwent surgery with general or regional anesthesia and used a wearable pedometer device within the 30 days before surgery and/or the 30 days after surgery. The primary objective was to assess whether pedometer data in the perioperative period were associated with postoperative outcomes (including length of hospital stay, development of complications, and unplanned readmission to the hospital). The secondary objective was the data capture rate (defined as the percentage of patients for whom adequate pedometer data were captured for analysis).
Search Strategy
This study conformed to the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) and Enhancing Transparency in Reporting the Synthesis of Qualitative Research (ENTREQ) statements.10 The study protocol is registered with PROSPERO (ID: CRD42021288843). We used search terms including “pedometer,” “wearable sensors,” “perioperative,” and their Boolean combinations in PubMed, EMBASE (Ovid), Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Web of Science Citation Index. We also hand-searched the bibliography of included studies for additional studies. The full search protocol is included in Supplemental Digital Content 1, Supplemental Material 1, page 1, https://links.lww.com/AA/E174. No language restrictions were imposed at the time of the literature search. All searches were conducted independently by at least 2 authors (Z.J., C.L., or K.Z.), and discrepancies were discussed after the search process.
Study Selection
Prospective or retrospective studies of adult patients undergoing surgery under general or regional anesthesia were included. The intervention of interest was the use of pedometer devices within 30 days before and after surgery. Only studies that reported the primary outcome in association with a low and a high perioperative step count cohort (as defined by the individual studies) were included in the analysis.
Data Extraction
Data extraction was done according to standardized proforma using Microsoft Excel. A second author then checked the data. Extracted data included bibliographical information, study design, and outcomes. Due to the expected heterogeneity in the timing of pedometer usage, pedometer data interpretation, as well as outcomes reporting, we did not plan for quantitative analysis.
We used the Risk of Bias in NonRandomized Studies of Interventions (ROBINS-I) tool, a 7-item questionnaire. Each item represents a risk category and can be determined to be low, intermediate (some concerns), and high, and this is then summarized as an overall grading.11
;)
Figure 1.: Search flow chart. CINAHL indicates Cumulative Index to Nursing and Allied Health Literature.
Figure 2.: Risk of bias summary. Green = low risk, amber = some concerns, and red = high risk.
Potential bias was evaluated at study and outcome level, and all assessments were done by at least 2 authors (Z.J., C.L., and K.Z.) independently. Any disagreements were discussed and resolved with a third author (T.J.G.). We used GRADEpro Guideline Development Tool (GRADEpro GDT, McMaster University, 2015) to assess the certainty of the conclusion that could be drawn from the available evidence.
RESULTS
The search was last updated on August 16, 2022. We screened a total of 1412 studies and identified 18 studies for inclusion (Figure 1). There were 2 randomized controlled trials that compared the masked pedometer to uncovered pedometer on postoperative ambulation, and the rest were cohort studies in which all participants were given a pedometer. Seven of the studies recorded preoperative pedometer data, while 13 studies recorded postoperative pedometer data. Notably, 10 of the studies were conducted on oncologic surgery patients. Interestingly, in-hospital data capture rates (63% to 94%) were not better than outpatient data capture rates (86% to 94%). The difference in data acquisition also does not appear to be due to patient age nor length of monitoring. Characteristics of individual studies are summarized in the Table 1. The risk of bias assessments of each study is displayed in Figure 2 and discussed in Supplemental Digital Content 1, Supplemental Material 1, pages 3–11, https://links.lww.com/AA/E174.
Preoperative Step Count
Preoperatively, Billé et al13 found that the first 2 quartiles walked 7673 and 4196 steps per day respectively, while the bottom 2 quartiles recorded an average step count of 967 and 2672 steps. There was no association between preoperative step count and cancer stage, performance status, or comorbidities. In a multivariable analysis, preoperative step count was an independent predictor of postoperative complications (odds ratio, 0.7 for each quartile; P = .001). The median length of stay in the top 2 quartiles was 3 and 4 days, respectively, while the median length of stay in the bottom 2 quartiles was 6 days; however, this was not statistically significant (P = .36). Risk of bias was intermediate.
Hedrick et al17 and Nakajima et al22 both stratified patients with <5000 steps per day as inactive. Hedrick et al17 reported that after adjustment for the National Surgical Quality Improvement Program (NSQIP) predictive risk scores, inactive patients had a significantly higher risk of postoperative complications (pneumonia, surgical site infection, and venous thromboembolism). While the length of stay showed a trend of association with the activity level, this was not significant (P = .88). Similarly, Nakajima et al22 reported that after adjusting for patient demographics, cancer, and surgery-related parameters, preoperative step count remained a significant predictor for postoperative complications (P = .042). Both studies had intermediate risks of bias.
Richards et al25 used a lower activity threshold of 2500 steps per day. After adjustment for age, tumor stage, and surgery, those who were inactive had a higher incidence of complications, which required surgical intervention (29% vs 9%; P = .04). Inactive patients also had a longer length of stay in the hospital (median 14 days vs 6; P < .01), and ICU stay (median 2 days vs 0; P < .01), and were less likely to be living independently at home 90 days after surgery (65% vs 90%; P = .02). The risk of bias was low to intermediate.
Two other studies by Esteban et al16 and Khetrapal et al19 did not implement specific activity threshold, but found correlation between preoperative step count and risk of complications after surgery. Esteban et al16 reported that after surgery, patients who had complications demonstrated a significant reduction in their daily step count (P < .001), which did not return to baseline, whereas those without complications recovered back to baseline within 4 days. Both studies had high risk of bias.
In summary, pedometer readings 1 day before surgery16 to more than a month before surgery17,22 correlated significantly with postoperative complication rates, but not significantly with the length of hospital stay. The overall certainty of evidence is low due to risk of bias across studies, heterogeneity in study design, as well as differing activity thresholds.
Postoperative Step Count and In-Hospital Outcomes
There were 2 clinical trials that randomized patients to receive an accessible pedometer or a blinded device during the postoperative period.24,27 Reed et al24 noted that the step count on postoperative day 0 significantly correlated with the length of hospital stay (P < .001). Wolk et al27 reported that step count on postoperative day 1 showed significant negative correlation with the length of hospital stay (P = .002); not reaching the median cumulative step count (4000 steps in open surgery and 5600 steps in laparoscopic surgery) was also associated with a significantly higher risk of complications and longer hospital stay. Interestingly, both studies reported that simply providing the patient with pedometer reading did not result in increased ambulation. Risks of bias were high and intermediate (Wolk) and high (Reed), respectively.
Two observational studies found that when adjusted for confounding factors, such as age and comorbidity, patients who ambulated more during the early postoperative period had a significantly shorter length of stay. Daskivich et al14 reported that on postoperative day 1, every 100 steps less than 1000 was associated with a 3.7% higher risk of prolonged stay (defined as length of stay >70th percentile for the given surgical procedure; P = .003). Nevo et al23 reported that patients who walked <1050 steps on day 2 had, on average, 3.4 days longer hospital stay (P < .01). There was also a significant correlation between postoperative step counts and postoperative complications (Spearman’s ρ = 0.39; P < .01 and ρ = 0.40; P < .05, respectively, for postoperative day 1 and postoperative day 2). Risks of bias were intermediate.
Other observational studies found that mean daily step count for the duration of the hospital stay may be negatively correlated with the length of hospital stay;20,29 however, the significance of this is unclear. Risk of bias was high. In a prospective cohort study of gastric resection patients, the authors reported that each 1000 extra steps on postoperative day 5 were associated with 2.72× increased odds of early discharge (P = .02) when controlled for the surgical approach and cancer types.28 Risk of bias was high.
While the included studies found associations between postoperative step count, length of stay, and in-hospital complications, the certainty of evidence is very low due to heterogeneity and data loss in some of the studies. It is also plausible that patients who developed overt complications may go through a period of subclinical deterioration that negatively affects ambulation. Lack of clear temporal separation between the timing of step count data and the development of complications makes the mechanism for this association difficult to establish.
Postoperative Step Count and Postdischarge Outcomes
Three studies found that lower postoperative step count during the hospital stay was associated with higher risk of unplanned readmission. Low et al21 noted postoperative in-hospital daily step count significantly correlated to 30-day readmission rate (odds ratio, 0.83 for each additional 100 steps per day; P = .01) and 60-day readmission rate (odds ratio, 0.82; P < .01) after abdominal cancer resection. Bae et al12 showed that after surgical resection for metastatic abdominal cancer, patients who were readmitted to the hospital had a significantly lower average daily step count during their hospital stay (652 vs 1299 steps; P < .001). Kane et al18 reported that patients who required readmission had significantly lower step counts on the day of discharge (964 fewer steps; P = .003) compared to the preoperative baseline.
Engel et al15 found that patients who developed postdischarge complications up to 6 weeks after cesarean delivery (including surgical site infection, postpartum hemorrhage, and endometritis) had significantly lower total step counts during their hospital stay (8719 vs 6974 steps; P = .04), but there was no significant difference between postoperative step count and in-hospital complications (8612 vs 7554 steps; P = .09). Incidentally, sleep length and quality were also measured; these did not demonstrate a significant association with postoperative complication rate. Risk of bias was intermediate.
Rossi et al26 conducted a reanalysis of pedometer data using machine learning to generate a predictive model for postdischarge complications or need for readmission using pedometer data and patient-reported outcomes (pain, fatigue, sleep, etc). They reported that a model generated from postoperative pedometer data could predict readmission or postdischarge complications, and the area under the curve of the receiver operating characteristics (AUROC) was 0.72 (meaning that the event assignment is correct in 72% of the cases). Additionally, patients with a low postoperative step count compared to preoperative levels, and those with a sudden increase in activity levels, also had higher risks of complications and readmission, which was attributed to exercise-related injury. The risk of bias was high.
The 5 included studies consistently reported correlation between lower postoperative step count during hospital stay and significantly higher incidence of complications or readmissions. The overall certainty of evidence is very low due to significant risk of bias across studies and heterogeneity in study design.
DISCUSSION
Pedometers, in various forms, are readily available and feasible to use in the perioperative period. In this systematic review, we identified a total of 18 studies that evaluated the link between perioperative pedometer reading and postoperative outcomes. Most studies reported significant correlation between preoperative step count and postoperative complication rate. Studies comparing postoperative step count and postdischarge outcomes found that patients who had lower postoperative step counts during their hospital stay had a significantly higher incidence of complications and readmissions. Although there is a somewhat limited volume of high-quality data, there may be potential for more global application of these devices to assess perioperative risk and outcomes. More specific data are required to establish the application of pedometers in day-to-day practice, in specific patient populations, and to corroborate the utility of these smart devices compared to commonly used current approaches to risk stratification or identification of frailty or fitness for surgery or postoperative discharge.
Paluch et al30 reported in a meta-analysis that among community-dwelling adults, there was a negative correlation between the daily step count and all-cause mortality. For patients younger than 60 years of age, the decrease in mortality plateaus at 8000–10,000; for patients older than 60 years, the correlation plateaus between 6000 and 8000 steps. The concept of a “ceiling effect” in perioperative ambulation is not well explored. Extrapolation from nonsurgical patient cohort overlooks the surgery-related pathology that limits ambulation, and the role of cardiopulmonary/functional reserve (as measured through ambulation) in postoperative morbidity versus outpatient all-cause mortality. In the postoperative period, ambulation goal is likely dynamic and reflects the varied functional recovery among different surgery and patient groups.
Pedometer reading is a measure of activity level rather than true physiological reserve. Several studies have demonstrated correlation between daily step count and CPET results. In a prospective study of patients undergoing CPET for major abdominal surgeries, daily step count 7 days before surgery moderately correlated with anerobic threshold, maximum oxygen consumption per min (VO2max), and peak work (Pearson’s correlation coefficients were 0.59, 0.39, and 0.48, respectively). Predictive models utilizing a combination of activity tracking and an international physical activity questionnaire demonstrate even better prediction of the CPET parameters.8 Before this, Novoa et al9 conducted a similar study in patients undergoing lobectomy and reported a moderate correlation between mean daily step count and VO2max. In a nonsurgical patient cohort without a diagnosis of myocardial infarction or heart failure, daily step count was shown to correlate with NT-proBNP level.31 In another nonsurgical patient cohort with heart failure, daily step count again demonstrated a strong correlation with VO2max.32 Watanabe et al33 conducted a study of >3600 older patients evaluating pedometer performance and the incidence of frailty. The authors reported that for every increment of 1000 steps per day, the odds of having frailty was reduced by 15% to 26%. Similar findings were reported by Lefferts et al,34 who reported that the odds of frailty was reduced by 38% per increment of 1000 steps per day. Pedometer data are currently not validated as a perioperative risk stratification tool; but larger-scale validation studies are justified, considering the correlation among pedometer data, postoperative outcomes, and CPET/NT-proBNP.
The review has several limitations. First, there was considerable heterogeneity in study design (timing of pedometer recording, surgical types, and outcome measures); therefore, it is not possible to quantitatively analyze the data, which increases the uncertainty of the evidence. Second, as most of the studies are nonrandomized in nature, the risk of bias was significant compared to randomized clinical trials. While studies adjusted for some clinical variables through multivariable analysis, the adjustment process is heterogeneous between the studies, and other measures of preoperative fitness were not considered. Finally, while the findings of the studies are suggestive of an association between a perioperative and postoperative step count, it does not establish causality. It is worth noting that there is a viable time course relationship between the step count and postoperative outcomes (Figure 3); the association was consistent across studies and is biologically plausible.
Figure 3.: Evidence of perioperative step counts and postoperative outcomes. Procedures in bold represent surgery types with multiple supporting studies.
It should also be noted that the percentage of patients with complete data sets varied significantly between the studies (Table 1). It was noticeable that smaller studies demonstrated a higher degree of variability in data acquisition, whereas most studies of >100 patients reported data acquisition in 85% to 90% of the patients. Other explanations for the varied compliance include patient motivation, health literacy, availability of technical support, institutional setting, and cultural factors. To put into context, it is not uncommon for intervention compliance to fall below 80% even in established enhanced recovery programs.35
Table 1. -
Characteristics of Included Studies
| Study ID |
Surgery |
Total patient number |
Median patient age |
Setting and duration of pedometer use |
Outcomes |
| Bae et al 2016
12
|
Abdominal cancer surgeries |
30 |
Not specified |
Postoperative in-hospital (excluding ICU time, median LOS 13 d) |
Readmission rate |
| Billé et al 2021
13
|
Lung cancer surgery |
90 |
66–73 |
Preoperative for 15 d |
Length of hospital stay, 30-d postdischarge cardiac and respiratory complications |
| Daskivich et al 2019
14
|
Metastatic abdominal cancer surgeries |
115 |
53 |
Postoperative in-hospital (median LOS 4 d) |
Length of stay |
| Engel et al 2021
15
|
Cesarean delivery under spinal |
218 |
32–34 |
Postoperative in-hospital (mean LOS 4.1 d) |
In-hospital complications and 6-wk postdischarge complications |
| Esteban et al 2017
16
|
Lung cancer surgeries |
34 |
64 |
1 d before surgery; postoperative in-hospital for 4 d |
4-mo postdischarge complications, and cardiopulmonary complications |
| Hedrick et al 2020
17
and Kane et al 2020
18
|
Colorectal surgeries |
106 |
51–60 |
30 d before surgery and 30 d after surgery |
In-hospital postoperative complications, length of stay, and 30-d unplanned readmissions |
| Khetrapal et al 2020
19
|
Radical cystectomy |
57 |
Not specified |
Preoperative for 7 d |
90-d all complications, Clavien–Dindo III, and above complications |
| Kizlcik Özkan et al 2022
20
|
Thorascopic lung resection |
52 |
60 |
Postoperative in hospital for 4 d |
Length of stay, pulmonary function |
| Low et al 2018
21
|
Abdominal cancer surgeries |
71 |
57 |
Postoperative in-hospital (excluding ICU time), mean LOS 12 d |
30-d and 60-d unplanned readmission |
| Nakajima et al 2020
22
|
Hepato-pancreato-biliary malignancy surgeries |
105 |
53 |
Preoperatively between the office evaluation and the day of surgery, median length of monitoring was 33 d |
In-hospital postoperative complication with Clavien–Dindo grade III or higher, in-hospital postoperative infectious complications, and length of hospital stay |
| Nevo et al 2021
23
|
Elective abdominal surgeries |
100 |
55
|
Postoperative in-hospital (mean LOS 2 d) |
Length of hospital stay and in-hospital complications |
| Reed et al 2021
24
|
Bariatric surgeries |
266 |
44 |
Postoperative day 0 to day 1 |
Length of stay |
| Richards et al 2020
25
|
Colorectal surgery |
85 |
76 |
Preoperative for 14 d |
Length of stay, complications, and 90-d postoperative disposition |
| Rossi et al 2021
26
|
Lung and gastrointestinal cancer surgeries |
52 |
Not specified |
Preoperative for 3–14 d, in-hospital, and 2–4 wk postdischarge |
30-d postdischarge complications and unplanned readmissions |
| Wolk et al 2019
27
|
Laparoscopic procedures |
132 |
59 |
Postoperative in-hospital for 5 d (mean LOS 15 d) |
Compliance with postoperative ambulation, length of stay, and in-hospital complications |
| Wu et al 2019
28
|
Gastric cancer resection surgeries |
43 |
68 |
Postoperative in-hospital and postdischarge for a total of 4 wk |
Length of stay (<9 d is defined as early discharge) |
| Yi et al 2021
29
|
Colorectal surgery |
59 |
39 |
Postoperative in-hospital (median LOS 5 d) |
Length of hospital stay, in-hospital postoperative complications, and Clavien–Dindo grade II or IIIa complications |
Abbreviations: ICU, intensive care unit; LOS, length of stay.
Current evidence suggests that perioperative step counts correlate with postoperative outcomes. It is not clear whether this is modifiable through increasing patient ambulation; nor is it clear how much ambulation and for how long would sufficiently reduce patients’ perioperative risks. As discovered by McDermott et al36 in the LITE (Low InTensity Exercise) trial, high-intensity walking exercise, but not low-intensity exercise, significantly improved 6-minute walk distance. This suggests that walk speed, as well as step count, may be clinically relevant in the perioperative period. As with prehabilitation programs, promoting patient ambulation through setting a targeted step count could improve patients’ functional status. It has the benefit of being much easier to administer than dedicated exercise programs and requires considerably less training for both health care providers and patients. Preoperative optimization is often a strong impetus for patients to engage in healthy behavior. As perioperative physicians, this represents an opportunity for behavior change that may lead to long-term benefits. Unlike CPET and NT-proBNP, pedometer reading is something patients can easily understand and self-monitor, thus it lowers the potential barriers for patient engagement.
CONCLUSIONS
This systematic review that investigated the correlation between perioperative pedometer readings and postoperative outcomes consistently demonstrated that both preoperative and in-hospital pedometer readings correlated with postoperative outcomes. Considering the widespread uses of pedometer devices, and the potential clinical value of perioperative pedometer monitoring, larger studies to validate the use of perioperative pedometer data for risk stratification are warranted.
DISCLOSURES
Name: Zhaosheng Jin, MBBS.
Contribution: This author conceived the topic and contributed to the literature search and writing of the manuscript.
Name: Christopher Lee, MD.
Contribution: This author contributed to the literature search, data extraction, and writing of the manuscript.
Name: Kalissa Zhang, MD.
Contribution: This author contributed to the literature search, data extraction, and writing of the manuscript.
Name: Rosen Jeong, BS.
Contribution: This author contributed to the literature search and review of the manuscript.
Name: Tong J. Gan, MD, MBA, MHS, FRCA.
Contribution: This author contributed to the review and editing of the manuscript.
Name: Deborah C. Richman, MB, ChB, FFA(SA).
Contribution: This author contributed to the writing, review, and editing of the manuscript.
This manuscript was handled by: Richard C. Prielipp, MD.
REFERENCES
1. Gupta H, Gupta PK, Schuller D, et al. Development and validation of a risk calculator for predicting postoperative pneumonia. Mayo Clin Proc. 2013;88:1241–1249.
2. Gupta PK, Gupta H, Sundaram A, et al. Development and validation of a risk calculator for prediction of cardiac risk after surgery. Circulation. 2011;124:381–387.
3. Tatematsu N, Park M, Tanaka E, Sakai Y, Tsuboyama T. Association between physical activity and postoperative complications after esophagectomy for cancer: a prospective observational study. Asian Pac J Cancer Prev. 2013;14:47–51.
4. Shinall MC Jr, Arya S, Youk A, et al. Association of preoperative patient frailty and operative stress with postoperative mortality. JAMA Surg. 2020;155:e194620.
5. Keeratichananont W, Thanadetsuntorn C, Keeratichananont S. Value of preoperative 6-minute walk test for predicting postoperative pulmonary complications. Ther Adv Respir Dis. 2016;10:18–25.
6. Hernandes NA, Wouters EF, Meijer K, Annegarn J, Pitta F, Spruit MA. Reproducibility of 6-minute walking test in patients with COPD. Eur Respir J. 2011;38:261–267.
7. Older PO, Levett DZH. Cardiopulmonary exercise testing and surgery. Ann Am Thorac Soc. 2017;14(suppl 1):S74–S83.
8. Jones L, Tan L, Carey-Jones S, et al. Can wearable technology be used to approximate cardiopulmonary exercise testing metrics? Perioper Med (Lond). 2021;10:9.
9. Novoa NM, Varela G, Jiménez MF, Ramos J. Value of the average basal daily walked distance measured using a pedometer to predict maximum oxygen consumption per minute in patients undergoing lung resection. Article. Eur J Cardiothorac Surg. 2011;39:756–762.
10. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097.
11. Sterne JA, Hernán MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919.
12. Bae S, Dey AK, Low CA. Using Passively Collected Sedentary Behavior to Predict Hospital Readmission | Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing. Association for Computing Machinery, 2016:616–621.
13. Billé A, Buxton J, Viviano A, et al. Preoperative physical activity predicts surgical outcomes following lung cancer resection. Integr Cancer Ther. 2021;20:1534735420975853.
14. Daskivich TJ, Houman J, Lopez M, et al. Association of wearable activity monitors with assessment of daily ambulation and length of stay among patients undergoing major surgery. JAMA Netw Open. 2019;2:e187673.
15. Engel O, Haikin Herzberger E, Yagur Y, et al. Walking to a better future? Postoperative ambulation after cesarean delivery and complications: a prospective study. Int J Gynaecol Obstet. 2021;157:391–396.
16. Esteban PA, Hernández N, Novoa NM, Varela G. Evaluating patients’ walking capacity during hospitalization for lung cancer resection. Interact Cardiovasc Thorac Surg. 2017;25:268–271.
17. Hedrick TL, Hassinger TE, Myers E, et al. Wearable technology in the perioperative period: predicting risk of postoperative complications in patients undergoing elective colorectal surgery. Dis Colon Rectum. 2020;63:538–544.
18. Kane WJ, Hassinger TE, Charles A, et al. Wearable technology in predicting readmission following colorectal surgery. Conference Abstract. Dis Colon Rectum. 2020;63:e128.
19. Khetrapal P, Catto J, Norman W, et al. The role of wearable devices and cpet in predicting major complications after radical cystectomy. Conference abstract. J Urol. 2020;203:e937.
20. Kizilcik Özkan Z, Gokce Isikil A, Ünver S, Yanik F. The effect of postoperative mobilization with a pedometer on pulmonary functions and length of postoperative hospital stay in patients undergoing lung resection: cross-sectional study. Arch Lung. 2022;21:15–23.
21. Low CA, Bovbjerg DH, Ahrendt S, et al. Fitbit step counts during inpatient recovery from cancer surgery as a predictor of readmission. Ann Behav Med. 2018;52:88–92.
22. Nakajima H, Yokoyama Y, Inoue T, et al. How many steps per day are necessary to prevent postoperative complications following hepato-pancreato-biliary surgeries for malignancy? Ann Surg Oncol. 2020;27:1387–1397.
23. Nevo Y, Shaltiel T, Constantini N, et al. Activity tracking after surgery: does it correlate with postoperative complications? Am Surg. 2021;88:226–232.
24. Reed B, Tabone LE, Tabone JK, Szoka N, Abunnaja S, Bailey K. The use of an activity tracker to objectively measure inpatient activity after bariatric surgery. Surg Obes Relat Dis. 2021;17:90–95.
25. Richards SJG, Jerram PM, Brett C, Falloon M, Frizelle FA. The association between low pre-operative step count and adverse post-operative outcomes in older patients undergoing colorectal cancer surgery. Perioper Med (Lond). 2020;9:20.
26. Rossi LA, Melstrom LG, Fong Y, Sun V. Predicting post-discharge cancer surgery complications via telemonitoring of patient-reported outcomes and patient-generated health data. J Surg Oncol. 2021;123:1345–1352.
27. Wolk S, Linke S, Bogner A, et al. Use of activity tracking in major visceral surgery-the enhanced perioperative mobilization trial: a randomized controlled trial. J Gastrointest Surg. 2019;23:1218–1226.
28. Wu JM, Ho TW, Chang YT, et al. Wearable-based mobile health app in gastric cancer patients for postoperative physical activity monitoring: focus group study. JMIR Mhealth Uhealth. 2019;7:e11989.
29. Yi Y, Sossenheimer PH, Erondu AI, et al. Using wearable biosensors to predict length of stay for patients with IBD after bowel surgery. Dig Dis Sci. 2021;67:844–853.
30. Paluch AE, Bajpai S, Bassett DR, et al.; Steps for Health Collaborative. Daily steps and all-cause mortality: a meta-analysis of 15 international cohorts. Lancet Public Health. 2022;7:e219–e228.
31. Parsons TJ, Sartini C, Welsh P, et al. Objectively measured physical activity and cardiac biomarkers: a cross sectional population based study in older men. Int J Cardiol. 2018;254:322–327.
32. Silver M. Can a pedometer serve as a surrogate for key exercise prognostic markers in heart failure? CHEST. 2015;148:20A.
33. Watanabe D, Yoshida T, Watanabe Y, Yamada Y, Kimura M, Group KS. Objectively measured daily step counts and prevalence of frailty in 3,616 older adults. J Am Geriatr Soc. 2020;68:2310–2318.
34. Lefferts EC, Bakker EA, Carbone S, Lavie CJ, Lee DC. Associations of total and aerobic steps with the prevalence and incidence of frailty in older adults with hypertension. Prog Cardiovasc Dis. 2021;67:18–25.
35. Li L, Jin J, Min S, Liu D, Liu L. Compliance with the enhanced recovery after surgery protocol and prognosis after colorectal cancer surgery: a prospective cohort study. Oncotarget. 2017;8:53531–53541.
36. McDermott MM, Spring B, Tian L, et al. Effect of low-intensity vs high-intensity home-based walking exercise on walk distance in patients with peripheral artery disease: the LITE randomized clinical trial. JAMA. 2021;325:1266–1276.
