gms | German Medical Science

Background: The role of e-Health in healthcare is growing rapidly. Also in back pain treatment e-Health programs have been shown to be effective. However, these programs can only unfold their positive effects if patients actually use them.

Question and objective: Which factors predict whether and how long patients use a web portal on back pain they receive from their general practitioner (GP)? We examined the predictive power of fifteen factors pertaining to patients’ demography, internet-related beliefs and behaviours, health and back pain, the GP-consultation in which patients received the portal as well as main targets of the web portal (e.g. knowledge about back pain).

Method: Prior using the web portal, patients received questionnaires that assessed the investigated factors. After a usage period of four weeks, patients completed a follow-up questionnaire that assessed whether they had used the portal. The web portal automatically recorded usage times of patients. Based on questionnaires from N=180 patients, we ran a multilevel binary logistic regression to examine which factors predict patients’ portal usage. Based on usage times of N=83 patients (85% of all patients who reported having used the portal), we ran a multilevel regression to examine which factors predict the length of usage.

Results: Regarding the chances of using the portal, four predictors approached significance (p<.10): The chances were higher for patients whose GP had used the portal in the consultation and for patients with chronic back pain. They were also higher for patients with poorer general health and for those who had less back pain recently. Regarding the length of usage, five predictors were significant (p<.05), two approached significance (p<.10): The portal was used longer by women than by men, longer by older patients than by younger ones and when patients had more back pain recently. It was also used longer the more satisfied patients were with the consultation (p=.06) and the lower their perceived efficacy in patient-physician interactions. Patients with a longer history of back pain and the more they believe rest is beneficial for back pain (p=.07) used it shorter.

Discussion: While the choice to use the portal at all was affected by patients' current condition (i.e., health status and back pain) and the GPs' behaviour, the length of use was affected by demographic characteristics and patients' experiences and beliefs about back pain and patient-physician interactions. The latter finding implies that patients with problems in patient-physician interactions might particularly benefit from our web portal as they seem to make more use of it and otherwise have limited opportunities to obtain good information. Interestingly, more pain seems to prevent patients from using the portal at all, but it also seems to motivate those patients who chose to use it to make more use of it.

Practical implications: The results show how GPs can support the usage of a web portal on back pain and illuminate which patients are particularly inclined to use such a web portal.

Appeal for practice in one sentence: GPs can support the usage of a web portal on back pain by showing the portal in the consultation and by addressing how rest and how using the portal might affect patients’ back pain.