Design and setting

We conducted a cross sectional online survey among primary care physicians in Hong Kong between October 2013 and March 2014. Ethics approval for the study was obtained from the Survey and Behavioral Research Ethics Committee (SBREC), the Chinese University of Hong Kong. The study followed the Declaration of Helsinki.

Doctors eligible for the survey were members of the Hong Kong College of Family Physicians (HKCFP, N = 1638). An online questionnaire with a cover letter was sent via email to all members registered with the HKCPF. One point of Continuous Medical Education (CME) score was awarded for every questionnaire submitted. A reminder was sent to the non-respondents 4 weeks later. We defined chronic low pain as pain or discomfort between the costal margins and the gluteal folds, with or without leg pain, which lasts for more than 12 weeks [2].

Measurements

The Pain Attitudes and Beliefs Scale for Physiotherapist (PABS-PT) was originally developed for physiotherapists, but more recently has been applied to a cohort of general practitioners in the Netherland, the UK, Brazil and Ireland [20–23]. It is a self reported measure that discriminates between a biomedical and a biopsychosocial orientation regarding the management of chronic low back pain. The scale was developed by expert validation and analysis of 36 items extracted from four different health-related questionnaires in 2003 [14]. The original 20 item PABS-PT was further validated by Houben et al. in 2005 [20], and this resulted in a 19 item, highly-rated tool for the assessment of health care providers’ attitudes and beliefs [24]. Each item is scored on a six point Likert scale that ranges from totally disagree (score 1) to totally agree (score 6). A higher score on each subscale indicates a stronger biomedical or behavior treatment orientation, respectively. A cut off point signifying a high or low score has not been established by the developers of the questionnaires. Gender, age, years of practice, type of practice (private or public), postgraduate training, clinical interests and personal experience of LBP were collected.

The participants were divided into four subgroups for further analysis using the scatter plot with biomedical (BM) score versus biopsychosocial (BPS) score. The four groups include (i) Group I: physicians with a low BM score and alow BPS score (ii) Group II: physicians with a low BM score and a high BPS score (iii) Group III: physicians with a high BM and a low BPS score and (iV) Group IV: physicians with a high BM and a high BPS score.

Statistical analysis

We hypothesized that Hong Kong (HK) practitioners have higher mean scores of BM than practitioners in Western countries, who, on average have a mean score of 30 and SD is 6.0 [13, 21]. A sample size calculation based on this hypothesis indicated a sample of 119 respondents was required to detect a difference of 1.8 with type I error set at 0.05, power of 0.9 and Cohen’s to be 0.3.

We also hypothesized that physicians with family medicine training would be more likely to fall inot Group II; whereas physicians without any family medicine training would more likely to fall into Group III.

The mean score and standard deviation of each subscale were calculated according to the methods specified by the questionnaire developers [14]. A Pearson’s correlation coefficient was calculated between the two subscales to evaluate any interrelationship. Descriptive statistics were used to summarize the demographic data of the participants. Multiple linear regression and multinomial logistic regression were used to investigate the demographic characteristics that could be associated with the beliefs and attitudes measured in the scale.

Main findings

This is the first survey of low back pain related attitudes and beliefs among a group of primary care physicians in Asia. The scores on the two subscales in the present study are very close. This may indicate an uncertainty among physicians regarding their treatment orientation [22].

Both BM and BPS scores are higher than those of a similar study conducted in the UK [21], which assessed the same domains with the same questionnaire items. This may be explained by the change in the undergraduate medical curriculum in our locality, which has involved the incorporation of more psychosocial elements into the traditional biomedical focused education [25, 26]. A direct comparison to the UK study with statistical analysis is not possible here, as we did not know the distribution of the data in the UK study.

Family medicine specialists have a lower biomedical score than General practitioners. They are not, however, necessarily associated with a high BPS score (group I and II). This may imply that current family medicine training may be good enough to prevent physicians from practicing with a predominantly biomedical approach. Whether a psychological model will be adopted may depend on multiple factors, such as time constraints, the practice setting and physicians’ own perceptions.

A strong contrast is noted between group I (low BM- low BPS) and group IV (high BM- high BPS). Physicians working in the public sector tended to fall into Group I; whereas physicians working in private sector tended to fall into Group IV. This can be explained by the fact that physicians in the public sector tend to minimize over-investigations and avoid abuse of public resources [27]. Therefore, they are more likely to adopt a less biomedical approach. Moreover, insufficent consultation time in a busy public setting and inadequate rapport with patients may potentially hinder them from adopting the biopsychosocial model. The opposite is true for private physicians, who are potentially more susceptible to medicolegal lawsuits and tend to over-investigate. They desire a medical diagnosis despite their understanding that psychosocial factors play a significant role in the pathogenesis of LBP [28].

Physicians in group IV are more likely to have mixed interests than those in group I. This is not surprising as physicians with knowledge and training in different aspects of LBP, starting from its pathogenesis to embracing the notion of psychosocial influences, will weigh both models as equally important.

Previous analysis showed that a biomedical and a biopsychosocial orientation were not two opposites on the same scale but, rather, that both factors were important in determining a physician’s orientation [14]. Until now, there have been no clear cut offs determined for “high” and “low” score on both BM and BPS subscales. There is also no evidence on the best combination of treatment orientation, although it is assumed that a low BM score and a high BPS score is preferred [15, 17]. Nevertheless, an individual physician’s worldview and orientation to the pain model will eventually influence his or her management decision on chronic LBP [12].

Limitations

There are several limitations to the present study: Firstly, the low response rate in this survey (9.52%,) can lead to a high sampling bias and potentially limits the generalization of the study results. A “non-response” survey will be valuable in minimizing this bias. Secondly, the survey was completed via an online system which potentially excluded those who were not proficient computer users. Thirdly, the survey recruited only physicians who are members of the HKCFP, which, once again, limited its generalization. Fourthly, the study only looks into the treatment orientations of primary care physicians; it may not reflect their true clinical behaviors. Finally, the multiple comparisons conducted in this study are associated with a false positive correlation with 5% of the significance by chance. Meanwhile, there is still no universally accepted approach for dealing with the problem of multiple comparisons; it is an area of active research, into both the mathematical details and the broader epistemological questions.

Implications for future research

The management of chronic LBP is moving towards a more integrated approach, especially in Asian countries, where Traditional Chinese Medicine and Acuputure are commonly practiced. Therefore, future studies can focus on comparing the attitudes, beliefs and clinical management of primary care physicians, orthopedic surgeons and Chinese Medicine Practitioners, with the concomitant evaluation of patients’ satisfaction. A new scope of these and how they relate to outcomes of patients is needed to generate the best model of care and to inform development of future implementation strategies.