The specific aims were a) to evaluate the agreement between the recommendations of the IP and the SP; b) to examine the effects of the SO on the patients' decision and the certainty of the decision; c) to evaluate the patients' satisfaction with the SO concept; and d) to examine the association between the recommendations of the SP and the main criteria for KA in the German AWMF guideline indications for KA (11).
This prospective cohort study evaluated an SO pilot project in patients with recommendations for KA by their IP. It was approved by the institutional review board at the medical faculty of the Ludwig Maximilian University Munich (project number 17–098). All patients signed informed consent forms prior to enrolment. The study was conducted in accordance with the Declaration of Helsinki. The analysis plan was registered in the open science framework before the analyses were performed (https://osf.io/).
The study was conducted at the Department of Orthopaedics, Physical Medicine and Rehabilitation (OPMR), University Hospital, LMU, Munich. OPMR is certified by Endocert as an endoprosthesis centre of maximum care. Endocert is the world's first joint arthroplasty-specific quality assurance system for certifying the quality of knee and hip arthroplasty. It had certified 543 in German facilities by the end of 2018 (18).
Patient recruitment and inclusion criteria
Patients were informed about the SO project by the AOK Bayern webpage, the AOK Bayern Facebook page, articles in the AOK members' magazine and by AOK Bayern branches. With over 3.5 million members, AOK Bayern has a market share of more than 40% among the statutory health insurance funds in Bavaria (19).
Interested patients called the OPMR. The patient flow, criteria for participation in the SO programme and additional criteria for inclusion in this study are shown in Figure 1.
The patients presented in person to the SP. The SP was a specialist in orthopaedics with at least 5 years of experience in KA. The SP evaluated the indication for KA by taking a medical history, performing a clinical examination, evaluating X-ray images and, if available, evaluating other medical reports.
In the medical history, the SP asked particularly about pain under exertion, at rest and at night, restrictions in daily life, quality of life, pharmacological and non-pharmacological treatments, previous surgery, subjective suffering, previous illnesses, psycho-social stress situations, contraindications for KA and risk factors for surgery. The SP examined the mobility and stability of the knee joint, crepitus, leg axis, clinical signs of inflammation and pain upon pressure of the knee joint structures. Depending on the medical history and symptoms, additional examinations were added. The SP assessed the radiological severity of osteoarthritis according to the Kellgren and Lawrence (K-L) scale (see section measures) (20, 21).
At the end of the presentation, the SP discussed the recommendation with the patient, taking into account the strength of the recommendation, the chances and risks of KA and other treatment options. In recommending KA, the SP aimed to follow the criteria of the German guideline indication for KA (11). The SP provided the patients with a report that included information on the results of the clinical examination, the evaluation of imaging, and the treatment recommendation.
T0: Participants received a set of questionnaires on the day of the appointment and completed the questionnaires before the doctor visit. During the visit, height and weight were measured, and X-ray images were evaluated.
T1: After the doctor visit, patients completed another questionnaire and submitted it on the same day.
The patient decision at T0 and T1 was evaluated by the question "Do you want surgical treatment for your knee?" Patients answered on a 5-point Likert scale: "Yes, definitely", "rather yes", "undecided", "rather no" and "No, definitely not". Decision confidence was assessed by the question "On a scale from 0 (not at all) to 10 (extremely), how confident are you about your decision for surgery?” These two questions were adapted from a longer decision quality instrument and translated into German. The answer options of the first questions were expanded from a 3-point to a 5-point Likert scale (22).
Pharmacological and non-pharmacological conservative treatment, previous surgeries, comorbidity and sociodemographic data were collected by closed questions.
The Knee injury and Osteoarthritis Outcome Score (KOOS) was used to measure knee pain (9 items), other knee symptoms (7 items), restrictions in activities of daily living (17 items), restrictions in sport and recreation function (5 items) and knee-related quality of life (4 items) (23–25). Each item was scored on a 4-point Likert scale. The scale ranged from 0 (worst) to 100 (best). For ease of interpretation of the regression model, the original scale was reversed prior to analysis (0 = best; 100 = worst). The KOOS has been validated in many languages, including German, and has demonstrated good reliability and responsiveness in patients with knee osteoarthritis and KA. Compared to the older, frequently used Western and Ontario Osteoarthritis Index (WOMAC), the KOOS has the advantage of fewer ceiling effects (25).
Generic health status was assessed by the 5-level version of the EuroQOL Group 5-Dimension Self-Report Questionnaire (EQ-5D-5L) (26, 27). It comprises five items that evaluate five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression (23, 24). An algorithm was used to calculate the EQ-5D index. The range was from − 0.661 (worst health) to 1 (best health). We also applied the EQ-5D-visual analogue scale (EQ-5D-VAS) (0 = worst; 100 = best). The EQ-5D-5L has been validated in patients with knee osteoarthritis and KA (27).
Anxiety and depression were assessed by the patient Health Questionnaire 4 (PHQ-4). The PHQ-4 is an ultra-brief, reliable and valid instrument with two items regarding anxiety and depression. Each item has four answer options (score 0–3). The scales range from 0 to 6. Scores ≥ 3 are considered probable cases of anxiety or depression.
In the course of the study, we added a question on the urgency of the recommendation of the IP because patients reported considerable differences in medical history. Accordingly, a more detailed description of the recommendation was considered important for a better understanding of differences between the IP and the SP.
The radiological severity of osteoarthritis was assessed by the K-L scale (20). The K-L scale is a commonly used system that classifies the radiological severity of osteoarthritis from 0 to 4 depending on joint space narrowing, osteophytes, sclerosis, and joint deformity of bone ends. Patients with grade 3 or 4 osteoarthritis show larger effects after KA than patients with lower grades (21).
At T1, patients were asked about the recommendation of the SP: "Did your second opinion doctor recommend KA?” The answer options were yes (surgery recommended immediately or within less than 3 months), no or “later surgery recommended depending on the course of the disease”.
To assess the influence of the SO on the patient's decision, we asked the following question: "How strongly does the second opinion influence your decision for or against a knee prosthesis?” The patients responded on a 5-level Likert scale (min = 1, max = 5): very strongly/strongly/somewhat/little/very little.
Satisfaction with the SO project was measured by the question: "What school grade do you give the AOK-LMU second opinion project?” The response options were adapted from the grading system of German schools: 1 = very good; 2 = good; 3 = satisfying; 4 = sufficient; 5 = deficient; 6 = insufficient.
Details of the statistical analysis plan were made publicly available before analyses were performed (https://osf.io/). The person who conducted the analysis (second author) was not involved in the data collection and was not employed at the SO department. Means and standard deviations were calculated for metric variables. Frequencies are expressed as percentages. Changes in the frequencies of decision preferences to T0 and T1 were tested for significance using the Chi-square test, and changes in decision confidence were tested for significance using the t-test for paired samples.
The association between the recommendation of the SP and the indication criteria of the German S2k guideline indications for KA was first analysed by descriptively comparing these criteria between patients with and without a current recommendation for KA. Then, the influence of these criteria on the recommendation was analysed in a multivariable logistic regression model. The independent variables were the K-L scale after transformation into a binary scale (1–2 versus 3–4), the KOOS scales for knee pain and knee-joint-related quality of life and previous treatments with exercise therapy (yes/no) or pain medication (no/on demand/always). The cut-off value for the K-L scale was set between 2 and 3 because patients with grade > 2 osteoarthritis show larger treatment effects (21). The dependent variable was the current recommendation for KA (yes/no). We did not use statistical selection of criteria because the model was designed to compare all of these variables. The Hosmer-Lemeshow test was used as the goodness-of-fit test.
Statistical analyses were performed with the software package IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.