This study is the only national study specifically to report use of CAM practitioners by older Australians who have chronic illness. Overall, 8.8% of our NSA sample, most closely representative of the wider population of older Australians, reported seeing a CAM practitioner in the previous three months. In common with other studies, we found that those who consult CAM practitioners are likely to be women and to have higher levels of education. However, in contrast with other studies [12, 17], we did not find that living in an area of higher socioeconomic status was related to higher use. We have also been able to show a differentiation in use with a decline in use with increasing age, with those over 80 years of age being the lowest users. In contrast to the Adams study  where 3.1% of those over 65 years visited CAM practitioners over the past year, our NSA sample showed use by 10.1% of those aged between 60 and 69 years; and 7.2% of those aged between 70 and 79 years in the past 3 months. It was only in the group of those aged 80 and above that the rate of use fell to 2.6%.
The different methodologies used by researchers in this area make direct comparisons of rates of use difficult. Many of the studies cited report CAM practitioner use over the previous twelve months (8, 17,) rather than the 3 months used in this study and use different age groupings and include different CAM practitioners. However, the proportions of those consulting CAM practitioners in the previous 3 months, at 8.8%, is at least as great as those reported by Adams  and more in line with the 26.5% rate reported by MacLennan and colleagues .
The gender difference in use is wider in our study (almost 3 F:1 M) than in other studies that report a 2:1 difference  but the reasons for this are not clear.
Studies looking at specific chronic conditions have found similar CAM usage to that in our study. In our study, 8.4% of respondents with recent treatment for cancer and 8.8% of respondents with a heart condition in the NSA sample reported accessing CAM practitioners. We found higher use of CAM practitioners among people with musculoskeletal conditions such as arthritis (9.2%), osteoporosis (15.9%) and chronic pain (13.0%) than cancer and heart disease. These figures are higher than those reported by Adams et al.  (5.6%, 4.4% and 8.5% respectively), even when reporting a shorter period.
While Westert et al.  have shown that use of health services increases with increasing ill-health and numbers of co-morbid conditions, the current study found that this did not apply to the use of CAM practitioners: neither self-assessed health nor the reported increasing numbers of chronic conditions was significantly related to their reported use. This may suggest that the reasons for consulting a CAM practitioner are linked to the presence of particular conditions, so that the actual number of conditions is less important, or, as other authors have suggested that it is related to the patient ‘world view’ in making their choice of provider  especially when “conventional care is not relieving their symptoms” .
This possibility is supported by the fact that the two chronic condition samples (Diabetes and Lung), for which conventional medicine has clear guidelines and treatment models had low CAM use.
Finally, respondents living in remote areas were significantly less likely than those in inner regional or metropolitan areas to report using CAM practitioners, a finding that may be related to the limited number of CAM (and other) practitioners in Australia’s remote areas, but that contrasts with Adams’ 2003 findings .
There is no commonly used group of practitioners included under a definition of CAM practitioner, and our study left it open to respondents to include CAM practitioners according to their own views. There is no capacity in this study as a result to link particular respondent characteristics to particular practitioners. Additionally the survey did not capture whether respondents saw mainstream healthcare practitioners who have been trained in CAM .
Using recall as the basis for data may lead to inaccurate reporting, which we hoped to minimise by using a limited period, particularly in relation to the number of times respondents consulted different health practitioners.
Our survey did not ask respondents whether they sought care from a particular practitioner in connection with a particular chronic disease or other health issue . It is possible that people with diabetes, for example, were accessing CAM practitioners for management of other conditions or to improve overall health .
The response rate from each of the sample groups is lower than we had hoped, which limited the analysis of all variables, and inevitably leads to a concern that these results are not representative. On the other hand, responses were received from respondents nationally, which provides a strength to the results.
Implications for management of chronic illness
The vast majority of those who consulted CAM practitioners also consulted mainstream medical professionals, almost universally in the case of respondents from both the Diabetes group and the Lung group. Respondents with conditions that have well evidenced treatment regimes that are widely used, such as hypertension, diabetes and asthma, have the smallest percentage seeking advice or treatment from CAM practitioners. On the other hand, higher proportions of people with conditions that are complex, involve pain management, and whose effects may be intractable using conventional medicines report seeking assistance from CAM practitioners.
We suggest that these findings have particular implications for mainstream healthcare practitioners. Other studies have shown that the combination of multiple care practitioners, as well as multiple medicines affect the wellbeing of people with chronic conditions [27, 28]. For this reason it is important that mainstream healthcare practitioners be aware of the healthcare choices their clients are making.
Robinson & McGrail’s literature review of disclosure of CAM and CAM practitioner usage found that the disclosure rate of usage to mainstream healthcare practitioners may be as low as 23%, with several reviewed studies reporting non-disclosure rates of 60-70% . The reasons for this include, but are not limited to, patient perception that the physician does not value CAM, physician is viewed as ignorant of CAM, physician does not ask about CAM usage, patient forgets to mention CAM usage . The disclosure of CAM usage to mainstream healthcare practitioners is essential, as Kiat and colleagues have suggested, to patient outcomes .