Sample characteristics
Most of the included patients had CBP for more than 10 years and visited their doctor frequently because of back pain (please see Table 2). Median pain intensity (Von Korff scale) in the intervention group was 60 [44.25–71.5] and in the control group 65 [52–75] (please see Table 4). Main characteristics of patients showed no statistical difference between intervention and control group. An overview of baseline characteristics of included patients and physicians can be found in Table 2.
- patients
|
intervention (n = 14)
|
control (n = 17)
|
age median [range]
|
59 [36-83]
|
57 [21-81]
|
gender
|
men: 9, women: 5
|
men: 4, women: 13
|
back pain duration
|
< 1 year: 0
1 year-10 years: 2
>10 years: 12
|
< 1 year: 1
1 year-10 years: 8
>10 years: 8
|
number of doctor visits because of back pain in the past 12 months
|
< 4 times: 9
4-12 times: 5
> 12 times: 0
|
< 4 times: 3
4-12 times: 12
> 12 times: 2
|
b) physicians
|
intervention (n=4)
|
control (n=4)
|
age median [range]
|
39.5 [28-64]
|
53 [36-57]
|
gender
|
men: 3, women:1
|
men: 2, women: 2
|
duration of working in primary care in years
[interquartile range]
|
6.25 [1.375]
|
19 [11.25-25]
|
type of surgery
|
joint practice: 4
|
joint practice: 4
|
size of surgery
|
big: 4 (>1500 patients per year)
|
big 4 (>1500 patients per year
|
location of surgery
|
urban: 1, rural: 3
|
rural: 3, ng: 1
|
Table 2a: Baseline characteristics of participating patients. Table 2 b: Characteristics of participating physicians/surgeries.
Study procedures were prolonged
We recruited participants from March 2022 to July 2022. Nine physicians (three women and six men) were recruited who then recruited 37 patients (18 women and 14 men). One physician (11%) and 6 patients (16%) could not be randomised because of a failure to contact them again. Physicians took an average of 38.5 days to recruit patients. Initially, we had planned 21 days for recruitment of patients by physicians and had to extend the recruitment period and the number of physicians recruiting patients. Please see Table 3 for detailed data on study procedures. Four physicians and their 14 recruited patients were randomly assigned to the intervention group, 4 physicians and 17 patients to the control group. All of the participants completed the baseline survey (t0) and all of the 14 patients in the intervention group underwent an ExPA guided consultation. This took place in a median 35.5 days after consent to participate. The variation here was very large (between 6 to 87 days). All physicians, and 13 patients could be interviewed after the consultation, one patient refused to be interviewed because of pain exacerbation. Most of the interviews were completed via telephone (telephone: 11, personal: 1). At t1, a total of 27 patients had completed the survey, 3 patients of the intervention group (22%) and one patient of the control group (6%) could not be reached again. Importantly, we could collect data of all participants of the intervention group at t1 as the participant, who could not be interviewed, filled out the questionnaires and the participants, who did not fill out the questionnaires, could be interviewed. All of the 4 physicians in the intervention group completed the survey, data collection from the physicians in the control group was not necessary. Figure 2 provides an overview of the participant flow.
|
actual time
|
planned time
|
Duration of recruitment of 6 physicians
|
50 days
|
63 days
|
Duration of recruitment of all 9 physicians
|
123 days
|
we planned to recruit only 6 physicians
|
Duration of recruitment of the first patient by physicians
|
median: 28.5 days, range: 3-51
|
na
|
Duration of recruitment of all patients by physicians
|
median: 40.5
range: 6-66
|
21 days for each physician
|
Number of patients recruited per physician
|
3 physicians recruited 4 patients
5 physicians recruited 5 patients
|
5 patient per physician
|
Time until consultation took place
|
median: 35.5
range: 6-87
|
na
|
Table 3: Overview of data on study procedures. Physicians took much longer to recruit patients and recruited less patients. Therefore, we had to extend the recruitment period and had to recruit more physicians.
Considerable rate of missing data
The number of missing data in the validated questionnaires (SF-12, IPAQ, HADS, von Korff) was considerable. This prevented further analysis; especially in the case of PA (only 50% of the patients in the intervention group filled out the IPAQ correctly). However, in the interviews, participants consistently expressed that they had no difficulties with the questionnaires. Furthermore, they emphasised that the process was not time-consuming. One interviewee highlighted the challenge of assigning a specific numerical value to her pain. Additionally, another participant expressed a preference for electronic questionnaires.
Table 4
Outcomes assessed at T0 (baseline) and T1 (3 months after intervention). Numbers are presented as medians with interquartile ranges in square brackets. IPAQ is presented as total MET (metabolic equivalent) minutes. HADS: Hospital Anxiety and Depression Score, A: Anxiety-Subscale, D: Depression-Subscale, IPAQ: International Physical Activity Questionnaire, ng: not given, SF-12 MCS: Short Form 12 Health Survey, Mental Component Score, SF-12 PCS: Short Form 12 Health Survey, Physical Component Score.
|
T0
|
T1
|
Von Korff pain graduation
(0 (no pain)-
4 (high pain-related impairment)
|
Intervention: 3 [2.5-4], ng:2
|
Intervention: 2 [1.25-3], ng: 4
|
Control: 3 [3–4]
|
Control: 2 [1–3], ng: 1
|
Von Korff pain intensity
(0 (no pain) –
100 (highest imaginable pain)
|
Intervention: 60 [44.25–71.5]
|
Intervention: 56.66 [41.5–64], ng: 3
|
Control: 65 [52–75]
|
Control: 59.8 [47.25–69.75], ng:1
|
SF-12 PCS
(0 (lower) –
100 (higher physical related quality of life))
|
Intervention: 37 [28–42], ng: 1
|
Intervention: 33.7 [32.7-39.82], ng: 5
|
Control: 32.5 [28.25–41.25], ng: 3
|
Control: 33.16 [27.53–39.56], ng: 3
|
SF-12 MCS
(0 (lower) –
100 (higher mental health related quality of life))
|
Intervention: 37 [35–45], ng: 1
|
Intervention: 51.8 [36.73–55.04], ng: 5
|
Control: 46.5 [35-54.75], ng: 3
|
Control: 39.6 [34.84–54.76], ng: 3
|
IPAQ
total MET (metabolic equivalent) minutes
|
Intervention: 3150 [1916–11088], ng:7
|
Intervention: 5412 [3807–9279], ng:7
|
Control: 1386 (595.5-1777.5], ng: 8
|
Control: 1515 [1173–4448], ng:2
|
HADS-A
0 (lower anxiety)-
21 (stronger anxiety)
|
Intervention: 8 [5–11], ng:1
|
Intervention: 8 [4.5–11], ng: 4
|
Control: 9 [6–11]
|
Control: 9 [5-10.25], ng: 1
|
HADS-D
0 (lower depression)-21 (stronger depression)
|
Intervention: 7 [6-9.75]
|
Intervention: 2 [1.25-3], ng: 4
|
Control: 5 [4–8]
|
Control: 2 [1–3], ng: 1
|
Increased PA and decreased pain in intervention group
Patients were satisfied with the counselling with ExPa (median 7 on a scale 0–10) and felt positive (median 7 on a scale 0–10) and motivated (median 6.75 on a scale 0–10) after counselling. Patients partly expressed an improvement of pain and the majority would recommend the counselling with ExPa (median 8.5 on a scale 0–10). This was also confirmed by the results of the qualitative interviews: Some patients expressed a greater sense of acceptance of their symptoms.
“I felt more accepted, I must say, with my complaints. And I actually found that positive” (patient_07).
The discussion atmosphere was seen as “pleasant” (patient_25) or at least as “quite normal” (patient_36) by patients as well as physicians. The main message of the counselling was conveyed to the patients:
“So (…) drugs don't have an effect as good as exercise, that is what I remember” (patient_07).
Many patients reported at least small changes in their PA directly after the consultation. They for example, started to cover more distances on foot or signed in a course:
“And then we came to the conclusion that I would prefer to do a course at home. And she [physician] prescribed it for me. And I had it approved by my health insurance” (patient_07).
Please see Fig. 3A for overview of results regarding the view of patients on ExPa.
ExPa contributed to a positive motivating consultation
Physicians generally felt somewhat supported by ExPa (median 6 on a scale 0–10) and they felt using ExPa contributed to a positive patient-physician relationship (median 7.5 on a scale 0–10). Physicians experienced the consultation as being quite motivating for patients (median 7.59 on a scale 0–10). In the interviews, physicians stated that they found ExPa helpful to structure their counselling:
“So I found that quite good as a guideline” (physcian_01).
They took more time to talk about PA with their patients and learned new aspects about them. Please see Fig. 3B for an overview of physicians view on ExPa.
Mixed Feedback and Suggestions for Optimizing ExPa
Satisfaction with using ExPa again was mixed (median 5 on a scale 0–10) and some features were not used often (e.g. specific information for doctors on different PA interventions). Also, in the interviews, few patients and physicians said that they did not to see additional benefits from using ExPa. They found the conversation partly too short:
“So it was surprisingly short. And for me it was also rather disappointing, I must say. I actually already knew what I was told" (patient_03).
In some cases, the graphical representation was found to be unconvincing. Some patients noticed the physicians´ uncertainties on using ExPa correctly:
“So I think Dr (…) is still having a bit of trouble with the programme." (patient_23).
Sometimes physicians did not show them the monitor or they did not use the function to print handouts. Physicians criticized the fact that the counselling with ExPa was quite time consuming and difficulties to integrate it in the practice routine. As a solution, they suggested the implementation of ExPa in the practice software.
On the other hand GPs as well as patients expressed also additional topics, which should be integrated into ExPa:
“So let’s say the occupational area or psychosocial stress factors … these should at least appear in a list like this, so that they are discussed (physician_09).
Satisfaction with study procedures
The study participation in general and the effort to participate in the study was seen as positive by patients (median 7.59 on a scale 0–10, respectively median 8 on a scale 0–10) and physicians (median 8.5 on a scale 0–10, respectively median 8 on a scale 0–10). This was also demonstrated by the results of the qualitative interviews. Most of the patients were satisfied with the effort and time required for participating in the study. Physicians sometimes criticised that the recruitment of patients and organizing the appointments was time consuming.
Figure 3 provides an overview of views on ExPa and the study participation.
a) patients view on ExPa
|
b) physicians view on ExPa
|
Figure 3a: Boxplot on patients view on “ExPa”. Dots = individual data points, vertical line = median, box = interquartile range, horizontal line = whisker-line (1,5xinterquartile range). Figure 3b: Overview of physicians view on “ExPa” and study participation. Dots = individual data points, vertical line = median.