We read with interest the report by Zagol and Krasuski on motorized scooters. The investigators suggested that scooters can have detrimental long-term effects on cardiovascular risk and that their findings pertinent to risk should influence physician practice. We are concerned that the study’s methods do not support such a conclusion. Specific limitations include the following: (1) the use of a retrospective cohort study design with no attempt to account for other factors potentially affecting outcomes (e.g., by using statistical, case-matched, or historical control), (2) incomplete data on other treatment pertinent to the outcomes being studied, and (3) very low survey response rates. These limitations must be considered in interpreting the study findings.

A retrospective cohort study design does not permit attributions of causality. It is impossible to know whether observed changes were caused by the natural histories of patients’ diseases. One recent study of the natural histories of diabetes and hyperlipidemia in a tertiary care outpatient program showed changes very similar to those in this population over the course of a year or so of follow-up: 25% of patients had diabetes at baseline, and on average, diabetic control decreased substantially over time (e.g., rates of diabetes doubled and mean glucose levels increased), while 60% had hyperlipidemia and average lipid levels improved substantially over time. The lack of a comparator group undoubtedly magnifies the attributed risks.

The problems with using a retrospective cohort study design are compounded by differential reporting on the medical management of diabetes, hypertension, and hyperlipidemia and differential attribution of causality. Zagol and Krasuski attributed unfavorable changes in blood pressure and diabetes control to the scooters and attributed favorable changes in the control of hyperlipidemia to medical management. Various alternative explanations exist that were not considered (e.g., variable efficacy and/or adherence with treatment for the various conditions).

The survey response rate was only 28% (far too low to attribute results to the population as a whole). Some journals, such as JAMA , will publish reports on survey results only when response rates are ≥60%. Despite this low response rate, Zagol and Krasuski reported metabolic parameters for the entire group, and they were not reported for the survey respondents. Thus, the survey nonrespondents are disproportionately represented in the metabolic outcomes. A more definitive conclusion could have been drawn by analytically coupling metabolic parameters with survey results. For all we know, the survey respondents may have been the group benefiting most from scooters, using them to get out and become more active, while the nonrespondents were recalcitrant to medical and lifestyle interventions, not using the scooters, and having more disease progression.

The data on scooter use were self-reported, and the actual survey instrument was not provided. Scooter use 4 hours/day is a surprising result. Typically, scooters are used for mobility outside the home, and patients walk inside. Because of scooters’ wide turning radii, their use within most homes is difficult; furthermore, scooter seats are uncomfortable for prolonged use. Patients who require power mobility inside and outside usually use power wheelchairs with specialized seating. Even then, the amount of use is highly variable, for example, averaging only 2 hours/day in patients with severe neurologic disease. In our randomized trial of motorized scooters for ambulatory patients, the scooters were used almost daily, commonly to transition between locations, and steps per day did not decrease in scooter users.

We agree with Drs. Zagol and Krasuski that further investigation is needed to understand health and functional impacts of scooters and identify the types of patients who benefit from scooters. It would be unfortunate if physicians and third-party payers were biased by this very limited study against providing equipment that may substantially improve the quality of life.