The questionnaire was completed by 129 (77%) of SCA members. The respondents who recorded their name (n = 92), were representative of SCA with regards to gender, age and years in clinical practise (Additional File 3).
The initial question of whether chiropractic care can be used preventively was by 126 participants answered with "yes, almost always" (n = 60) or "yes, sometimes" (n = 66). Thus, 98% of the responding chiropractors seem to support the concept of MC.
The hypothesis that tertiary preventive care is recommended to patients only if they improve considerably, was supported by the results obtained at the initial workshop (Additional File 1). According to the participating chiropractors, the patient should improve at least 50% before the clinician would recommend preventive care. When a patient shows a 76–80% improvement, most groups would consider MC to be suitable.
The hypothesis that secondary preventive care is recommended to patients with a history of previous LBP was also supported by the workshop (Additional File 1). However, the result from the workshop suggested that several other factors were taken into consideration before making this decision, factors further explored in the questionnaire survey. A summary of the replies to the specific questions obtained in the survey can be seen in Additional File 4. Two factors were found to have "good agreement" as "very important", namely i) the frequency of LBP over the past year, and ii) the frequency of LBP over the past 10 years. Eight factors out of 14 were found to have "reasonable agreement" as "very important", namely duration, (over the past year and of the present attack), treatment (effect and durability), lifestyle, work conditions, psychosocial factors and patient attitude. One factor had "reasonable agreement" as "not important", namely the patient's ability to pay for the treatment. Respondents failed to answer 8 times (0.4%) in 7 different questions.
An "other" factor was listed by 28 (22%) of the participating chiropractors. Most suggestions (n = 12) mentioned patient motivation (e.g. "patients' priorities", "if the patient wants a better health", "if the patient is expecting MC"). Some (n = 3) considered patient compliance (e.g. "patient's ability to follow advice"), some (n = 9) examination findings (e.g. "neurological status", "palpable dysfunction", "posture"), and a few (n = 4) miscellaneous answers were noted (e.g. "age", "body awareness", "effect on organic problems"). The research team would have placed the 12 patient motivation replies under "patient attitude". However, upon further scrutiny, 11 of the 12 respondents noting motivation as important had already noted "patient attitude" as "very important", so recoding this would not have affected the results.
Eighteen of 20 (80%) returned the retest questionnaire (Additional File 5), but only seventeen questionnaires were valid. Agreement was calculated in several different ways, and results ranged from 60% (perfect agreement by categories) to 72% (defining agreement as less than 16 mm using the line as a VAS, measuring continuous data), which we consider to be acceptable reliability.