Preparing for an orthopedic consultation using an eHealth tool: a randomized controlled trial in patients with hip and knee osteoarthritis

Non-surgical treatments like lifestyle education, exercise therapy, weight loss and pain medication are recommended as a primary approach to manage hip or knee osteoarthritis (OA) in an early stage and can be organized in primary care [1, 2]. Once these conservative treatment options have been tried adequately and have failed, or in case of diagnostic uncertainty, referral to an orthopedic surgeon should be considered for further evaluation and consideration of surgical interventions, e.g. a total joint replacement (TJR) [2]. Among medical, economic and healthcare professionals’ factors, patient-related factors may influence the choice for a TJR [3]. To actively participate in the consideration of different treatment options, patients need to be informed of benefits and possible disadvantages of available treatment options [4].

Currently, half to two-third of patients referred to an orthopedic surgeon are considered not (yet) eligible for a TJR [5, 6]. This is in contrast with the observation that patients with hip or knee OA who are referred for a first orthopedic consultation often expect action to be taken [7], in particular the planning of a TJR. It is conceivable that expectations for some patients are not met, resulting in patients being dissatisfied [8].

An appropriate preparation of the consultation is likely to streamline the patients’ expectations and increase their satisfaction, irrespective of whether the outcome is consideration of surgery or not. This hypothesis is supported by the literature where, in general, patients who are more knowledgeable, skilled and proactive prior to a consultation are more satisfied with received care because it is more likely that needs are met [9]. Interventions aimed at supporting patients to prepare for a consultation were found to improve self-efficacy in older patients [10]. Moreover, educational tools have high satisfaction rates and positive effects on patient knowledge, decision making, self-efficacy and number of questions asked during consultation [11, 12].

An educational eHealth application may be suitable to prepare patients for their consultation because of the easy accessibility and the possibility to provide information that suits individual preferences and needs. Moreover, eHealth interventions have shown to enhance and supplement the communication between patients and healthcare providers and seem effective at providing information, enhancing information exchange, and promoting self-management in older adults [13, 14]. Recently it was concluded that the use of an educational website for patients with hip and knee OA improve important aspects of quality of care (i.e. self-management, lifestyle and physical activity) [15]. However, these results were based on an observational study and to our knowledge good quality randomized controlled trials evaluating educational eHealth tools with interactive parts are not available.

The aim of the present study was to evaluate the effect of a stand-alone mobile and web-based educational intervention (educational eHealth tool) compared to usual preparation of a first orthopedic consultation of patients with hip or knee OA on satisfaction with the consultation. Secondary outcomes were knowledge, treatment beliefs and measures on the consultation from the patient and surgeon’s perspective.

Between March 2017 and March 2018, 836 individuals were invited to participate. A total of 293 (35%) participants filled in the first questionnaire and were randomized. No differences were found between the invited patients who did not participate and the study population with regard to age (P-value = 0.08) and sex (P-value = 0.61). Due to time constraints inclusion was stopped after 293 out of the targeted 382 patients were enrolled. Data of 7 (2%) participants was excluded because they did not fulfil the inclusion criteria; 5 participants cancelled their appointment and 2 participants were wrongly included as it turned out they already had been to the clinic before for OA complaints in the same joint. Two hundred eighty-six participants were allocated to either the intervention (n = 144) or control (n = 142) group. Nineteen (7%) participants were lost to follow-up leaving data of 267 participants for the ITT analysis. Twenty-eight participants in the intervention group did not open the application and were considered protocol-violators and excluded in the per-protocol analysis (Fig. 2).

Baseline characteristics of the study population are shown in Table 1. The majority of patients was female (58%), around 80% of patients had a consultation with regard to complaints on the knee and more than half experienced their first symptoms in the previous 5 years.

The results of this study show that preparing a first orthopedics consultation for hip or knee OA using an educational eHealth tool does not result in higher satisfaction of the consultation. Participants using the eHealth tool did have more knowledge and less negative beliefs about physical activities and pain medication as compared to usual care. No other significant differences between intervention and control group were found regarding treatment strategy (surgery versus other strategies), or treatment beliefs about TJR. Neither were there differences between groups regarding the evaluation by the orthopedic surgeon on preparation and active participation of the consultation.

We hypothesized that patients are more satisfied with their consultation regardless of the chosen treatment when they are well prepared using an educational eHealth tool. The results do not confirm our hypothesis, but are in line with a recently published RCT demonstrating no improvement in the appreciation of the first orthopedic consultation, after the use of an educational smartphone app [28]. However, secondary analysis showed that patients with a TJR planned as result of the consultation were significantly more satisfied than patients who had a different outcome, irrespective of the use of the tool. This confirms findings from previous qualitative studies that patients want action to be taken [6] and prefer TJR as treatment [29]. Previous research shows that expectations of TJR are often high and not always realistic [30, 31], but are a key determinant of treatment satisfaction after TJR [32‐34]. Also, patients’ beliefs about conservative and surgical treatments options are an important aspect in the choice for a treatment [3, 35]. Our results are in line with recent findings showing that educational tools used either during or in preparation for the consultation improve knowledge and risk perception [28, 36]. However, effects of improvement in knowledge and risk perception on fulfilment of expectations and ultimately satisfaction still has to be investigated [37].

The lack of positive effects of the tool on satisfaction may have several explanations. First, it may indicate that our intervention was not comprehensive enough to sufficiently change patients’ expectations towards other treatment options in order to improve their satisfaction. Although fulfilment of expectations is an important aspect of satisfaction [8, 33], healthcare professional related aspects like trust, empathy, communication and relatedness, contact time and waiting time are important for satisfaction with the consultation as well [8]. These aspects were not specifically targeted in our intervention. Second, in hindsight our choice for using the CQI as outcome measure may be questioned for several reasons. Because of missing values we could not calculate indicator scores for all participants. Moreover, it should be noted that satisfaction was high in all patients. Scores found in our study were even higher than previously reported CQI scores in orthopedics setting (3.3 (hospital stay) and 3.5 (information at discharge)) [27]. Moreover, > 15% of participants scored the highest possible CQI scores indicating a ceiling effect on our primary outcome [38]. The primary focus of the eHealth tool was to target patient expectations. Fulfilment of expectations was not reflected in the CQI and as such not measured. A validated sensitive questionnaire to assess satisfaction incorporating fulfilment of patient expectations about the consultations and its outcomes is needed, but currently unavailable.

Based on the positive effects on knowledge and beliefs found in this study, further implementation of the educational eHealth tool may be valuable for clinical practice. More guidance on the use of our educational eHealth tool and instruction for the orthopedic surgeon to discuss preparing the consultation may result in better outcomes [39] and should be a focus of further implementation. However, an advantage of eHealth interventions is that they can be used irrespective of time and place, without involvement of a healthcare professional. If the ultimate aim is to routinely use the educational eHealth tool, costs and benefits with regard to what is effective and what is practical should be weighted and further studied. Additionally, our intervention is already suited for preparation for consultation with other healthcare professionals (e.g. physiotherapist or general practitioner). It may therefore also be implemented in primary care, where it is likely that beliefs and expectations regarding treatment strategy are initially formed and where use of the tool can contribute to providing consistent education throughout the treatment process for OA patients.

There are several limitations of this study that need to be addressed. First, we included fewer patients than intended which impacted statistical power. The number of patients visiting our clinic with (suspicion) of OA was lower than expected which resulted in a slower inclusion rate. Due to time constraints inclusion was closed after 293 out of 382 intended patients were enrolled. Although loss to follow-up rates were lower than expected (7 and 25%, respectively), the final number of participants of whom data could be analyzed was slightly lower (n = 267 instead of n = 286) than aimed for based on sample size calculation. Considering the small differences found, it is not likely that we failed to detect an effect that was present (Type II error). Second, because the entire study was web-based we created bias against eHealth illiterate participants and may have excluded a relevant group of patients [40]. We did not systematically examine reasons for not participating because of practical reasons. Although gender and age of responders did not differ significantly from non-responders, further research into factors associated with participating in eHealth studies and use of the educational eHealth tool could provide starting points for improvement of the application [10, 39]. Last, we did not ask participants to discuss the tool with the specialist, but we cannot rule out that participants brought up the intervention during their consultations. Therefore, the orthopedic surgeons were not blinded in this trial, which might have changed the specialists’ behavior, outshining the effect of the educational eHealth tool.

In this randomized controlled trial we demonstrated that an educational eHealth tool did not result in higher satisfaction with a first consultation in orthopedics outpatient clinic setting for patients with possible hip or knee OA. The eHealth tool did have small effects on knowledge and treatment beliefs. Future research is needed to evaluate if improving the educational eHealth tool and optimizing implementation in different care settings result in better outcomes.