Background
Postoperative recovery is an essential component of surgical therapy. Innovation of surgical techniques and the development of interventions aiming to enhance the recovery process, necessitate tools to measure recovery [
1‐
5]. A frequently used standard to measure postoperative recovery is length of hospital stay, because it provides us with a well-defined and objective definition [
6]. However, an increasing number of operations are performed as day surgeries and hospital stay has decreased generally over the last decade [
7]. Patients are not fully recovered at time of discharge and subsequently, hospital stay does not correlate with postoperative recovery. Currently, a wide variation of instruments and outcomes are used to determine postoperative recovery, such as quality of life, satisfaction, pain, recovery indices and return to normal activities [
1,
8‐
10]. Unfortunately, all of these measuring instruments are subjective and self-reported and therefore prone to measurement bias [
1].
In physical activity research, accelerometers are used as an objective measure of physical activity and sedentary behaviors [
11,
12]. Furthermore, activity monitors such as pedometers and the now widely commercially available wearables are used to promote physical activity, for example as behavior change tool in trying to prevent and manage obesity [
13,
14]. Promoting physical activity is an important element in medical care as well, for example in the prevention of postoperative complications such as pneumonia, decubitus and deep venous thrombosis [
15‐
17]. The use of accelerometers in postoperative care seems therefore logical, not only to measure postoperative physical activity levels, but also as a tool to stimulate physical activities after surgery.
Some earlier research has been carried out with accelerometers in perioperative care, but these studies have focused on a relatively short postoperative period or used the accelerometer as an intervention to stimulate physical activities [
18‐
20]. Therefore, data representing the normal recovery process measured with an accelerometer are lacking. In this observational pilot study we will measure physical activities during four moments of one week each in the postoperative period. We aim to evaluate:
1)
Whether recovery of physical activity levels can be measured postoperatively with the use of an accelerometer.
2)
Whether the physical activity results measured with the accelerometer correspond with self-reported recovery of physical activity levels.
3)
Whether the use of an accelerometer in the postoperative course is feasible and accepted by patients.
Discussion
Results of this proof of concept study show that the accelerometer provides a feasible and objective method to measure recovery of physical activity levels after various forms of abdominal surgery. The group of patients who underwent minor surgery reached their mean baseline step count and MVPA three weeks after surgery. The group of patients who underwent intermediate surgery had not reached their mean baseline step count and MVPA five weeks after surgery. However, determining whether or not each individual patient has reached his or her own baseline step count, shows that only 61.5% (8/13) of the patients who underwent minor surgery and 25% (3/12) of the patients who underwent intermediate surgery, had reached their baseline step count or MVPA at the end of the study (five weeks after surgery). An explanation for this dissimilarity is that eight patients had surpassed their baseline levels five weeks after surgery, which increased the mean group levels. The outcomes of the accelerometer showed a fair agreement with the self-reported activity results recorded in the activity diary and a moderate agreement with patients’ report of having the feeling of being physically recovered.
There are few other studies which use the accelerometer in the postoperative care. In a study by Bisgaard et al. [
18], 24 patients undergoing laparoscopic cholecystectomy used a wrist-worn mini-motion logger Actigraph accelerometer (Basic model, Ambulatory Monitoring, Inc, New York, NY) one week before surgery until one week after surgery [
18]. Patients reached their baseline physical activity level two days after surgery. The difference in outcome might be explained by differences in study population. The pre-operative activity level in our study was lower than the pre-operative activity level in Bisgaards’ study (105.07 activity counts/minute vs around 200 activity counts/minute), which suggests that the study population in Bisgaards’ study consisted of a more active group. However, we have to be careful comparing the activity levels from both studies, since different measuring instruments are used. Secondly, our group of patients who underwent minor surgery, comprised three different types of laparoscopic surgery, of which three out of 13 underwent a laparoscopic cholecystectomy. The heterogeneity of our group might result in different outcomes. Lastly, the study group of Bisgaard et al. followed a relatively intense study program which might have resulted in a difference of motivation.
Another study carried out with an accelerometer in postoperative care by Wasowicz-Kemps et al. including an intervention group (
n = 36) and a control group (
n = 28) which both wore an accelerometer (PAM Model AM101, PAM B.V., Doorwerth, The Netherlands), performed measurements one week before a laparoscopic cholecystectomy until one week after surgery [
19]. The control group received no feedback from their accelerometers as the display was turned off, while the intervention group did receive physical activity feedback from the display after surgery and also received personal advice regarding their physical activity. Activity scores were expressed in PAM scores based on accelerations in the vertical axis. In the control group 10/28 (36%) patients had reached their preoperative value one week after surgery compared to 18/36 (50%) patients in the intervention group. Reaching baseline was defined as a PAM score exceeding 90% of the mean preoperative value. In our study only 2/13 patients (15.4%) who underwent minor surgery reached their baseline step count one week after surgery and 5/13 (23.1%) their baseline MVPA. In our study the definition of reaching baseline was also set at 90% of the mean preoperative value. Again, the difference in outcome might be due to the heterogeneity of our minor surgery group and the fact that patients from the intervention group of Wasowicz-Kemps’ study received extra advice regarding their physical activities and could not see their own activity results.
Until now, our pilot study has been the first study in which postoperative physical activity levels are measured for a period longer than one week after surgery with an accelerometer, without any additional interventions, resulting in a clear and objective view of the recovery process. Only in obesity surgery, studies have been performed measuring activity levels after bariatric surgery nine months and 12 months after surgery respectively [
28,
29]. However, these studies are focusing on measuring physical behavior changes instead of measuring recovery of physical activities after surgery. Another strength of our study is that we also measured self-reported activity results, using a self-developed activity diary. We have chosen to do so since pre-existing questionnaires focusing on daily activities were not suitable to match to the accelerometer results, because we were interested in a date of resumption after surgery and this is not what these kind of questionnaires focus on. We tried to develop a well-documented list, which has not yet been validated in this population, by using a functional ability list which earlier had been used to develop the convalescence recommendations of the type of surgical procedures we included in our study [
24,
25]. However, since the agreement between the accelerometer results and the activity diary was fair in this study, no assumptions can be made regarding the fact that this was because we have not used a validated questionnaire to measure activities or because the correlation between self-report and objective measures was indeed low. This could be the case since the accelerometer only measures objective physical data and the self-report also contains subjective mental feelings regarding the recovery process. The same applies to the moderate agreement between patients’ self-report of being physically recovered and reaching baseline step count with the accelerometer. Future research in this field should therefore be performed using a validated physical activity instrument (such as the International Physical Activity Questionnaire [
30]) alongside the accelerometer. Another limitation of our pilot study is the heterogeneity within the small study sample with regard to surgical procedures. Different surgical procedures might result in different patterns of recovery. However, the heterogeneity could also be regarded as an advantage, as using an accelerometer to objectively assess postoperative recovery levels might be applicable to a wide range of surgeries. We have tried to remedy this by dividing the surgical procedures into minor and intermediate surgery. Although the subdivision is partly based on current literature, the subdivision remains controversial. Therefore, future studies should try to include more participants per surgical procedure in order to improve comparison between and within the different patient groups. Another limitation is that the mean wear time per day differed significantly between the measuring weeks. Wear time during T1 was lower, which was due to the alternative definition that was used for this time point as many patients were bedridden and not always wearing the accelerometer in bed in this week. This may give an underestimation of physical activity in the first postoperative measuring week, especially in patients that might not have been bedridden and did not wear the accelerometer. However, wear time in the first week after surgery was considerably lower than in the other weeks mostly because participants did not wear the accelerometer in bed, and hence a correction for wear time would overestimate the physical activity levels in the first week considerably as non-wear time was mostly spent sedentary. Overall the protocol adherence was good, which suggests that the underestimation will have been minimal. Wear time during T2 and T3 was also lower than during T0, however in our opinion this difference was clinically not relevant and possibly still due to extra sleep time. Lastly, there was a relatively low inclusion percentage. The non-response analysis showed no significant differences in patient characteristics between included and non-included patients, but the included patients are likely to be a selection of the more motivated patients who are more willing to accept the additional burden of wearing an accelerometer.
This study has shown that the accelerometer can be used as an objective tool to measure recovery of physical activity levels after laparoscopic abdominal surgery (hysterectomy, adnexal surgery, cholecystectomy and hernia inguinal surgery). Step count and MVPA showed the same postoperative pattern which suggests that measuring postoperative step counts is an adequate measure for assessing postoperative activity level. Although this study showed that it is possible to measure postoperative recovery by an accelerometer, the clinical application remains controversial. First of all, because of the fact that using the accelerometer is relatively time consuming. This was supported by the low inclusion percentage in this study and the fact that although all patients completed the study, four patients could not fulfilling the strict wearing time criteria. A small wrist-worn device might have the advantage over the current hip-worn device, being easier to wear during the bed bound period just after surgery, resulting in better wear compliance. In addition the accelerometer which we used in this study costs approximately EUR 250 per device. Researchers can consider using a cheaper commercially available activity tracker which only measures steps in future research or clinical practice as opposed to our more expensive research based tool. Further, we think that the relatively low inclusion percentage was also because we could not offer the participants any benefit since they did not receive any feedback in the current study. However, when an intervention will be added, ie when patients can see their own activity results and get feedback, the added value might become obvious. The results of this study can help defining some standard values for giving feedback. In conclusion, an accelerometer can be useful in postoperative care to measure postoperative physical activity levels. Future research with a larger number of patients, also including open abdominal surgical procedures have to be performed to get more insight into the generalizability and clinical applicability of the results of this proof of concept study.
Acknowledgements
Not applicable.