Procedure

Eleven Danish chiropractors were selected by a senior chiropractor with knowledge of their clinical practice profile. The inclusion criteria were that they used maintenance care and were willing to submit to the interview. The interview instrument (described below) was sent out per mail and replied to in a subsequent telephone interview, conducted by the first two authors, who also analyzed the data.

The structured interview

The same questionnaire as that used in the previous Swedish survey was employed [3]. It had been designed by the group of Swedish chiropractors, who were responsible for the first survey, and further refined in a series of small pilot tests, in which the intellectual quality and clarity of the questions was addressed. The questionnaire that was written in English consisted of a basic case, describing a hypothetical patient with low back pain.

The basic case was described as follows: "A 40-year old man who consults you for low back pain with no additional spinal or musculoskeletal problems, and with no other health problems. There are no aggravating factors at work or at home. His X-rays are normal for his age. There are no "red flags"."

This case was followed by a description of nine possible ways in which this case could develop ("scenarios") (Additional file 1). For each scenario the respondents could choose between six different clinical management strategies in response to the question: "What would you recommend?" In the questionnaire, these are listed under A-F. An additional, "neither of the above, please explain"- alternative was also available (G), with the possibility to provide ones own alternative answer (Additional file 2).

Replies to the questionnaire were noted by the interviewers as categorical answers, i.e. strategy A, B, or C etc. Suggested additional strategies were noted down. If a new suggested alternative (G) was, in fact, identical with one of the previously suggested alternatives, it was placed under the correct alternative. Combinations of replies were coded as "several replies" and only truly original suggestions could be left under alternative "G", i.e. "None of the above, please explain".

The interpretation of all replies was done independently by the first two authors. Answers on which both agreed were considered factual. In cases of disagreement, the third author would be called in.

The open-ended question: "What is your definition of the concept of maintenance care?"

In addition, the chiropractors were encouraged to provide their own definition of maintenance care as response to an open-ended question ("What is your definition of the concept of maintenance care?"). These interviews were taped, transcribed and analyzed. Already during the interviews it became clear that the chiropractors generally failed to provide a clear definition of maintenance care but that they rather described various aspects associated with it. For this reason, we had to approach the analysis in a different way by specifically looking for statements that related to 6 pre hoc defined headings relating to: 1) LBP history, 2) type of symptoms, 3) severity of symptoms, 4) specific objective findings, 5) frequency of treatments, and 6) duration of maintenance care program. Statements that did not fit under any of these headings would be listed separately with the intent of finding additional descriptors under which they could be classified. Explanations relating to the definition of maintenance care were thereafter placed under the relevant pre- and post hoc defined headings. This analysis was done independently by the first two authors. Items on which both agreed were considered factual. Cases of disagreement would be discussed with the third author. Finally, the frequency of statements under each heading was quantified and entered into a table.

Proportion of patients treated with maintenance care

At the interview, the participants were asked to identify the number of patients receiving maintenance care out of their total number of patient visits, on the day of the interview. From this information it was possible to calculate the percentage of maintenance care patients seen by each chiropractor, and the mean and median values for the whole group.

Ethics

The participants were known to the first two authors, and the interviews were recorded, later to be transcribed, with the permission of the interviewees. To protect their anonymity, no names were included in the transcribed document and tapes were destroyed as soon as the written text was available thus cutting the link between the participants and their responses. No permission from the local ethics committee is needed for this type of project.

The use of maintenance care according to the structured interview – results of the Danish study

No problems in understanding or relating to the questions of the survey were noted among the respondents. Also the analysis was unproblematic.

As can be seen in Table 1, maintenance care would be recommended by at least some of the respondents for all but the last two scenarios. Both symptom-guided and clinical findings-guided maintenance care were noted for the remaining scenarios.

Some type of maintenance care was the most common answer for scenarios 2, 4 and 5. The largest degree of consensus was noted for scenario 4, which described quick and full recovery in a patient with recurrent symptoms. "Symptom-guided maintenance care" was the preferred answer for 72% whereas the remaining 28% would recommend "clinical findings-guided maintenance care", i.e. 100% would recommend maintenance care.

In all 54% would recommend some type of maintenance care for scenario 2; a worried patient who asks for maintenance care but who had only 2 days' of LBP and no previous history. For scenario 5, which is described as a patient with several short-lasting events over the past year but with no obvious improvement after 6 visits, some type of maintenance care was also considered relevant by 55% of the respondents. However, the single most common response for scenario 5 was "try again".

On the other hand, none of the participants suggested any form of maintenance care for scenario 8 (patient getting gradually worse) or scenario 9 (fluctuating symptoms with signs of co-morbidity). Instead, the largest groups (55% and 45%, respectively) agreed that such patients should be submitted to a "second opinion".

More than half of the group agreed on the "quick fix" approach in scenario 1 (immediate recovery and no previous low back pain).

Almost 2/3 would "try again" with the patient in scenario 3 (unexpected lack of improvement in seemingly uncomplicated case). "Try again" was the preferred management strategy also in scenario 6 (initial substantial improvement followed by stagnation), and in scenario 7 (initial substantial improvement followed by gradual worsening).

Strategy D ("external help – keep in touch") or strategy G ("other") were never the most often selected responses.

Open-ended question: "What is your definition of the concept of maintenance care?"

Only one of the chiropractors expressively defined maintenance care, using the terms of primary, secondary and tertiary prevention, stating that he/she would use maintenance care synonymously with secondary and tertiary prevention only, and. not as a means to prevent the incidence of low back pain (primary prevention).

In order of the frequency with which the information was provided, maintenance care can be described as follows: The aim is to prevent relapses by improving spinal function, and the consultation is a "check-up". Thus, it is irrelevant whether the patient has symptoms or not and the severity of symptoms is also not important. The frequency with which the patient is scheduled would vary with the needs of the patient, probably between 1 and 6 months, with 3 months being relatively common. An indication for maintenance care would be recurrent symptoms over a long period of time.

Interview vs. mail survey – validation of the questionnaire

In a short questionnaire, it can be difficult to convey meaningful clinical information to be understood by a wide variety of individuals. By using the same questionnaire as in the Swedish study but collecting the information through a structured interview, it would be possible to pick up any misunderstandings or comments that could shed light on problems of interpretation in the original questionnaire. Fortunately, the interview revealed no difficulties for the chiropractors to understand or relate to the questionnaire.

Open-ended question-results

The present study included also an open-ended question that was meant to shed some more light on the definition of maintenance care. However, with one exception, the chiropractors in the present study did not provide distinct definitions of maintenance care but approached the question indirectly, by describing some clinical aspects thereof.

In doing so, they did not describe any conditions or profiles or any standard management programs. Rather, their responses indicated that maintenance care can be used for any type of low back pain with the intent of improving spinal function in order to avoid a relapse of the condition. In fact, this corresponds to the definition of secondary prevention, and none of the respondents discussed a tertiary prevention scenario.

What people think that they do and what really happens, do not always concur. According to a recent Norwegian multi-center outcome study of patients with more persistent or chronic low back pain, at the one-year follow-up, maintenance care was shown to have been given mainly to patients who did not have a good short-term outcome [5]. In other words, absence of symptoms in the short term did not seem to incite further treatment (secondary prevention) but continued symptoms did (i.e. tertiary prevention). Whether this was the chiropractors' preferred choice or whether patients with quick recovery were uninterested in secondary prevention, is of course not known.

Some of the interviewed chiropractors considered it irrelevant, if the patient had symptoms or not, at the time of the consultation ("clinical-findings guided maintenance care"). If patients' symptoms are thought to be irrelevant, the chiropractor would have to trust his own examination findings entirely. This is a big responsibility, in the absence of evidence for the validity of the tests used to determine the presence of the treatable spinal lesion or without much knowledge of other factors that precipitate the recurrence of low back pain. Obviously, this concept requires further study.

Further, almost all of the interviewees were of the opinion that chiropractic treatment can improve spinal function and that optimal spinal function will reduce the risk for recurring problems of low back pain. This may be a commonly held belief. According to Rupert, it was commonly thought among North American chiropractors that one of the purposes of maintenance care is to "determine and treat subluxation" [6], confirmed in a similar study in Australia [7].

Judging from our results, there is no commonly agreed standard frequency of scheduled treatments but it would vary with the needs of the patient. This differs from the impression one gets when reading the case-report of one maintenance care patient by Wenban [8]. According to his report a pre-determined, hence probably rigid, treatment schedule was offered to the patient. We were unable to find other descriptions of this aspect in the literature but such a standardized approach appears unreasonable. How can one predict in advance the future clinical development and consequential treatment needs of an individual patient?