Flying Blind
The Australian Health Data Series

Flying Blind is a series of three reports dedicated to uncovering the acute levels of data fragmentation existing at all levels of Australia’s health landscape.

in collaboration with

The Ethics Quagmire:Case Study 2

Service Delivery or Research -  A Case of Consent

Health and medical researchers are sensitive to the privacy, confidentiality and security of the data they collect or are provided with. In addition, they are most aware of the need for ethics approval(s) to use or gather data to conduct research and analysis. Indeed, a comprehensive understanding of data security, informed consent and ethical use is central to the professional life of a health and medical researcher. But the types of projects, and their data requests and analysis, are as diverse as the persons undertaking the research.  It is vital then that our governance policies and processes for data, ethics, consent and privacy are able to respond to these differences. In this blog we look at one example where this has not been the case.


The type of data or data sources sought by researchers is dependent on the nature of the research questions.  Thereby, scholars in undertaking their investigations may decide to:

  1. Collect primary data through the use of surveys, particularly when the data required is not readily available, as in survey-based research projects, such as the 45 and Up Study;
  2. Collect primary data through the conduct of randomised control studies used to examine the impact of new interventions as in clinical trials;
  3. Collect primary observational data as in observation of during the course of routinely delivered healthcare, as in efficacy studies;
  4. Seek access to existing data sets that were routinely collected during the course of healthcare delivery, as in population-level studies;
  5. Seek to link existing administrative datasets from several jurisdictions to answer multi-disciplinary research questions (as in the work by Kathy Tannous);
  6. Seek to link service delivery data with private or public sector administrative data or medical and pharmaceutical claims data to analyse the impact of service delivery on overall healthcare utilisation.

These categories of research projects are not the same. They have varying risk profiles, methodologies, patient(s) contact and data requirements. It makes sense then that these differences would benefit from slightly diverse approaches when it comes to ethics approvals and obtaining consent. The case study for this blog is a project that falls under category 6 above.

Case Study Details

A primary care provider decides to enhance their existing care model through coordinated clinical care for high risk chronic disease patients. This type of service delivery model is in response to the federal government’s Health Care Homes initiative. In making this change, they plan to analyse the uptake and impact of their new coordinated care model as well as effects of this model on healthcare utilisation by the target group over time. This type of study is somewhat unusual as it is not a randomised clinical trial in the traditional sense, nor is it a purely observational study. This is due to the study seeking to understand the impact of the coordinated clinical care model on healthcare utilisation. To make their analysis more robust, the plan is to link this routine care delivery data with observational datasets from pre-existing hospital administrative data sets, as well as Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS) data. For the process of linking their delivery data, they are required to obtain additional consent from their patients, over and above the initial agreements to receive the enhanced care delivery model. The patients at medical risk will receive the care treatment irrespective of their consent to participate in the program or have their medical data linked.

The Challenge

This research project aims to use data linkage of primary care data with hospital administrative data, to gain more robust insights into the impact of their new service delivery model.  To get access to linked data in this way requires ethics approval that entails as the first step of getting appropriate patient consent to link their service level data with their existing health administrative data sets such as hospital data. In a service-delivery focussed project such as this one, getting multiple consents can become a very complicated process.  As service delivery happens across multiple jurisdictions and in different institutions, consent forms become complex, particularly when the names of the data sets used for linkage have to be explicitly specified in each consent form. Correspondingly, the required approvals also span multiple data custodians and ethics committees’ boundaries.  Even before the formal ethics application can be lodged, the researcher must spend a substantial amount of time designing multiple consent forms and related participant information sheets, mandated by different data custodians and ethics committees. This not only places a significant burden on clinicians and researchers, but is also incredibly confusing for patients. In this case, some of these patients had already consented to the care being delivered and to the linking of their de-identified data for research analysis. The same patients are then being asked to re-consent based on the consent forms structure or wordings specifications from the ethics committees or data custodians requirements.

One service delivery question; multiple consent requirements; hours of researchers' non-value added time


For the service providers delivering coordinated clinical care, obtaining consent from their clients using multiple complex-worded forms is a very difficult and costly process. In this ongoing multi-jurisdiction case study, procuring approval from multiple ethics agencies and data custodians for consent forms and participant information sheets has involved hours of work over two months. The notional dollar value of researchers’ time spent on consent forms alone amounts to several thousand dollars - all this to obtain what is categorised as ‘low risk de-identified administrative data sets’.

The bigger risk to this study is the potentially reduced uptake of much-needed coordinated primary care for an at-risk patient group. The irony being that the very spirit of designing consent forms to prevent ‘harm to humans’ is defeated by the heavy handed bureaucratic processes. Signing multiple complex consent forms can confuse participants who may choose not to consent to healthcare that they actually need to save them from harm.

So if this type of service delivery focussed study had to start only after receiving ethics approvals, a few high risk patients might not survive the wait! 

Although NHMRC have clearly articulated the many benefits of adopting a national approach to single ethics review of multi-centre research, in practice this is often not the case. 

Watch the space as we explore the national mutual acceptance scheme in a future blog.


Post new comment