Tracking Postoperative Recovery—Making a Case for Smartphone Technology
Menée aux Etats-Unis à partir de données portant sur 62 patients atteints d'un cancer traité par chirurgie, cette étude analyse l'intérêt d'utiliser les données fournies par l'accéléromètre de leur smartphone (évolution de l'activité physique quotidienne, capacité à faire au minimum 60 minutes d'activité physique par jour) pour évaluer leur capacité à récupérer après une opération chirurgicale, selon la survenue d'un événement indésirable
Recovery is a key concept in surgery, yet it is difficult to define or measure. Assessing postoperative recovery traditionally uses short-term surrogate measures, such as hospital length of stay and complications1; however, these outcomes fail to capture the perspective of patients who equate recovery with resuming usual activities after hospital discharge. It is thus essential that patient-centered recovery assessments incorporate data that evaluate physical function beyond the immediate postoperative period.
The use of smartphone activity sensors provides an opportunity to capture physical activity data after surgery, enabling the extended assessment of patients along their recovery trajectory. In a prospective cohort study in this issue of JAMA Surgery, Panda et al4 tested the hypothesis that metrics derived from a smartphone accelerometer would be able to detect differences in physical recovery in patients undergoing cancer surgery. Their findings support that, among patients with a postoperative event (ie, complication and/or reoperation), accelerometer data captured a decrease in daily activities up to 6 weeks after surgery. Similarly, fewer of these patients achieved 60 minutes of daily exertional activity compared with patients without a postoperative event. The methodological pitfalls of the study include a risk of selection bias (only 57% of eligible patients downloaded the smartphone application) and attrition bias (accelerometer data were missing for a median of 22% of days). Also, there is no mention of a study protocol setting hypotheses a priori; therefore, reporting bias cannot be excluded. The lack of a specific methodological framework for the validation of health-associated measures obtained from mobile devices is a current (major) limitation in this research field.
JAMA Surgery , commentaire en libre accès, 2018