IBD patients from three different care settings in two distinct geographical locations in Australia were invited to participate. This method has been reported previously[20].

The first cohort came from a metropolitan public teaching hospital based specialist IBD Service at Flinders Medical Centre (FMC). This is a large, government funded hospital, offering secondary/tertiary care for a local regional population of 341000 with a Gastroenterology inpatient and outpatient service, and IBD nurses available to patients within working hours.

The second cohort consisted of IBD patients in an overlapping area, receiving their care via a metropolitan Private Practice setting. These patients were under the care of one of four male general Gastroenterologists with extensive experience in managing IBD, without attachment to a specialist IBD unit, or access to IBD specialist nurse support.

The third cohort consisted of IBD patients cared for via Royal Darwin Hospital (RDH), a public hospital in a very remote location in Northern Australia. When this study was conducted, IBD care in Darwin was undertaken predominantly by general practitioners (GPs) and general surgeons, with no specialist gastroenterologist residing in Darwin, and no IBD nurse. The nearest tertiary hospital is in Adelaide, SA, more than 3000 kilometres away.

Potential subjects were identified from IBD databases/hospital records in each location and mailed a questionnaire. Reminder letters were sent to non-responders after one and three months.

The opening section of the questionnaire sought demographic details including age, gender, disease type, indigenous, relationship and employment status as well as current or previous history of smoking.

In the following sections, A-D, participants answered questions assessing: (1) views regarding conventional IBD medications; (2) views regarding CAM; (3) quality of Life; and (4) psychological and personality traits. Where possible, validated instruments were used as described below.

IBD-specific CAM use was assessed by asking subjects to rate the frequency with which they use complementary or alternative medicine to treat IBD on an ordinal Likert scale. A dichotomous variable was then generated whereby “yes” responses encompassed those describing their use as “often” or “very often”, and “no” included responses “sometimes”, “rarely” and “never”.

Medication Adherence was assessed using the Morisky 4 item Self Report Measure of Medication Taking Behaviour[21,22], examining predominantly dose omission, and covert dose reduction (CDR), the tendency to take less than prescribed of IBD medication without prescriber awareness was assessed as a dichotomous variable (yes/no) based on answer to the question “I take less than prescribed of my IBD medication without telling my doctor”. This has been previously reported[20].

Free text responses regarding attitudes towards IBD medication and dose modification were encouraged.

Other non-standardised attitudinal statements were put to subjects, seeking their views regarding IBD treatment beliefs and attitudes. Some Likert data were collapsed into categories “yes” and “no” for data presentation, but analysed as ordinal data or continuous data using factor scores for regression analysis.

Anxiety and Depression were measured using the Hospital Anxiety and Depression Scale[23], higher scores indicating higher levels of anxiety or depression. Quality of Life was measured using the reliable and valid Short Inflammatory Bowel Disease Questionnaire[24].

The Spielberger State-Trait Personality Inventory[25-27] was used to assess and compare depressive symptoms, anxiety, anger and curiosity between cohorts in both the immediate (state) and long term (trait or personality characteristic).

Comparisons between cohort means and medians were performed using the Kriskal Wallis test for non-normally distributed values, and two tailed t test or ANOVA for normally distributed values. Pearson’s χ² or Fisher’s exact test were applied as appropriate for categorical data.

Significant or trend associations at univariate level (P < 0.10) determined which variables were included in regression analyses, along with demographic factors.

Additional continuous variables summarising themes across the questionnaire were generated using principal component analysis for ordinal data using M Plus software (V5.2), for the purpose of data reduction. An oblique (oblimin) rotation was used of 37 of the 55 Likert scale items assessing all aspects of IBD treatment. An examination of the Kaiser-Meyer Olkin measure of sampling adequacy suggested the sample was favourable (KMO = 0.618). When loadings less than 0.4 were excluded, the analysis yielded an 8 factor solution. Scores for each of these 8 factors were normally distributed.

Binary logistic regression was used to assess predictors of CAM use as a dichotomous dependent variable, adjusting for age, gender, employment and relationship status.

A P value of < 0.05 was considered statistically significant. Apart from factor analysis, statistical calculations were performed using IBM SPSS Statistics for Windows, version 22, 2013 (IBM Corp). The statistical methods for this study were reviewed by Dr Reme Mountifield of Flinders Medical Centre, South Australia.

Response rates to the survey differed between cohorts, with 337/612 (55.1%) of FMC and 91/180 (50.5%) of SA private invitees participating, compared with 35/100 (35%) in Darwin (P < 0.0001). Non respondents did not differ from respondents by gender (P = 0.2), but there was a trend toward non respondents being younger than respondents (mean age 43.7 vs 50.3 years, P = 0.065) Darwin subjects were more likely be current or previous smokers, and to receive a disability support pension. This population has been previously reported[20]. Demographic data are summarised in Table 1.

Many subjects (45.4% overall) reported regular use of CAM, with no significant difference in usage frequency between cohorts (P = 0.594) (Figure 1). Distribution of CAM type used is presented in Table 2, and was not significantly different between cohorts (P = 0.626). The regular use of more than one CAM type (i.e., physical as well as homeopathic methods) was reported by 64.5% of subjects.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Proportion of subjects in each cohort reporting regular complementary medicine use.

Rates of CAM use were higher amongst younger (46.69 vs 53.41 years, P < 0.001), female (52.0% vs 35.5%, P < 0.001), and permanently employed (51.1% vs 37.4%, P = 0.004) subjects. However, CAM usage did not differ by disease type (P = 0.394), conventional medication pill burden (P = 0.784), smoking status (P = 0.805), or vegetarianism (P = 0.256) on univariate analysis.

Of the 206 subjects who reported regular CAM use, 52.5% felt that it was effective (worked “well” or “very well”), and 20.7% had obtained the therapy at consultation with an alternative practitioner rather than independently. The vast majority (83.3%) continued to use conventional IBD medications concurrently. Only half (54.1%) discussed their CAM treatment with their doctor, despite 87.6% of subjects reporting feeling comfortable doing so.

In contrast, of those reporting previous consultation with an alternative practitioner only 62.2% felt comfortable discussing conventional therapy with their alternative practitioner (P < 0.001), and 16.6% reported the CAM practitioner discouraged their use of conventional IBD medication. With regard to the consultation experience, 10.5% felt less intimidated by alternative practitioners than doctors, and 16.9% felt more informed about IBD by the alternative practitioner.

Of the 194/206 (94.2%) subjects who offered reasons for their CAM use, 33.0% reported safety concerns regarding conventional medications. Subjects who elaborated further expressed the belief that “natural” CAM would enable them to reduce reliance on “chemical” conventional therapy and dose reduce or cease these medications. Seeking a holistic approach to health in some way was cited by 32.0%, and 20.6% report advice from family, friends, colleagues, religious advisors, or the internet as their main reason for use. A smaller proportion (14.4%) cited lack of efficacy of conventional medications in treating IBD. No significant cohort based differences were observed.

Attitudinal and behavioural associations of CAM use on univariate analysis are presented in Table 3.

Of all subjects including CAM users and non-users, 57.3% reported family or friends using CAM for any health purpose. Those with CAM-using contacts was more likely to use it themselves for IBD (59.9% vs 40.1%, P = 0.004), free text responses suggesting that type of CAM chosen was influenced by social contacts.

The 54.9% of subjects reporting adverse effects of conventional medications were more likely to use CAM (P = 0.025), as were the 26.9% reporting regular self-initiated dose reduction of medication (P < 0.001). Lack of doctor communication satisfaction was reported by only a small proportion of patients (2.4%) but was associated with CAM use, as was seeking of psychological or psychiatric treatment (P < 0.001) when analysed as individual items.

Analysis of HADS, QOL and Spielberger mean scores suggested that increased anxiety, higher quality of life and lower depression scores were associated with increased CAM use, whilst personality type did not influence rate of use (Table 4).

After adjustment for age, gender, disease type and employment level, attitudinal and psychological predictors of regular CAM use using binary logistic regression analysis are shown in Table 5. This model explained a significant proportion of variance in low adherence rates (adjusted pseudo R squared 0.217, goodness of fit Hosmer Lemeshow P = 0.161).

After adjustment for demographics a trend was observed toward higher CAM usage amongst non-smokers (OR = 1.299, 95%CI: 0.993-1.698, P = 0.056).

Covert dose reduction, lower depression scores and subjects’ propensity to seek psychological help predicted CAM use, the latter factor analysis generated variable encompassing use of antidepressants, and consultations with counsellors, psychologists or psychiatrists (Table 5). Similarly, the factor analysis generated variable assessing dissatisfaction with doctor communication was an independent predictor of CAM use, and included satisfaction level with doctor relationship, doctor communication style, level of comfort in asking questions of doctor, and comprehension of information provided during consultation.

Complementary and alternative medicine (CAM) use is common in inflammatory bowel disease (IBD), and some demographic and clinical predictors have been established. This article reports a cross sectional survey of Australian subjects from contrasting care settings with IBD, focussing on the frequency and type of CAM use, and its behavioural and attitudinal predictors.

In the world of ever increasing influence on everyday health decisions from social contacts via social media and the internet, physician understanding of patient context needs to evolve to promote strong and open partnerships with patients in making treatment decisions.

This study demonstrates that patients with IBD make decisions regarding CAM use that are subject to multiple inputs, only one of those inputs being the treating physician. The importance of unqualified health advice from social contacts needs to be acknowledged and addressed in order to optimise adherence to conventional therapy.

IBD physicians need to attend more closely than ever to clear communication with patients regarding the risks and benefits of conventional therapy, and enquire about CAM use to better understand the patient’s context. Patient understanding of disease and therapy should be routinely assessed by physicians in order to correct misperceptions introduced by social contacts, alternative practitioners and the internet that may undermine successful IBD treatment.

This well presented study makes an important contribution to the literature as it highlights the broader context of individuals with IBD, reporting new behavioural predictors of complementary medicine uptake in this population which warrant attention during consultation.