Immersive Place-Based Attachments in Rural Australia: An Overview of an Allied Health Program and Its Outcomes

Round 1

Reviewer 1 Report

The focus of this paper is to understand the link between healthcare education and career choice - specifically choosing to work in underserved rural locations.  Given the lack of practitioners and the impact this has on both healthcare service and health itself, this makes this paper of great importance.  I enjoyed reading the paper as the complex data and analyses are presented lucidly and transparently - I really commend the authors on the effort they have put in here.

The findings are that rural background positively predicts future rural practice, an extracurricular factor.  There authors also present data that suggests that taking part in rural placements, also positively predicts rural practice, albeit to a lesser extent.  Furthermore, there is a 'dosing effect' - the longer the placement the bigger the effect of career choice.  This is very encouraging to the rural healthcare education field as such placements are an often expensive and logistically difficult to perform.  As such the authors conclude and additive effect of rural background and rural exposure.  

With the above in mind my major question to the authors is how strong these conclusion can be or whether they should be moderated? I say this because the program offer optional rural experiences, and from the data given it appears that those with rural backgrounds are more likely to choose rural placements.  As such, the two variables are likely not independent and hence caution is needed in interpreting the odds ratios, particularly the tone that the analysis conducted in Table 5 suggests a causative effect rather than merely correlative - indeed the entire design cannot prove causation.  I would suggest the authors add some 'hedges' to the their text, particularly in the abstract, as a different student body makeup of other programs's could mean they will see more or less impact of these sorts of placements on career choice.

Further the 'dosing effect' may be due to rural background students being more likely to choose the longer placements (Table 3).  How strongly do the author's trust the analysis to tease these different variables apart?  

The paper also has the interesting finding that urban background students benefit more attitudinally (to rural work) than rural background students (for example Table 3).  Does this mean there is some ceiling effect occurring and that perhaps we would get more educational benefit sending urban students to rural placements than the rural background students who, on the basis of this study, a substantial number of which practice rurally irrespective of what placements they do?  This is of importance as rural placements may be in short supply in some locales and knowing who gets access to produce the greatest impact becomes of importance.

I would say these points are discursive in nature and do not detract from what is a very interesting study which has the potential of being very impactful.

 

Author Response

Please see attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

A useful contribution to the literature especially as they have follow up data. of actual rural practice rather than just intention. The response rates at follow up are low but consistent with those from MDANZ surveys and reflective of busy graduates perhaps not placing a high priority on responding to research. 

I feel the methods could be improved by adding more information about the survey tools rather than just referencing them. 

The analysis could also indicate what other variables were associated with rural intention and practice in multivariate analysis. 

Given the smaller cohort in study undertook longer immersions and most undertook shorter ones - a clear power statement should be included 

Discussion needs comment on limitations especiaslly low response rate, cohort being disproportinately of shorter rather than the longer immersions, response bias with those more favourably inclined to rural practice perhaps being inclined to respond and those who did not feelijng disengagement and not responding. Some check of variables between responders and nonresponders in follow up might reassure about managing bias of responders themselves to ongoing participation.

It was unclear if external validation of responders follow up was made. Were AHPRA working postcodes sought? 

Overall a useful contribution and only minor changes suggested

Author Response

Please see the attachment

Author Response File: Author Response.pdf