Concord Hospital Concord Hospital
The Concord Research Office

Completing a HREA form for a medical records project

NOTES ON COMPLETING THE HREA FORM

FOR RESEARCH INVOLVING MEDICAL RECORD REVIEW

To download a sample HREA form, please click here. The below notes pertain to the relevant questions on the HREA form for a medical records project.


Q1.1 By appending the suffix “:A review of medical records” to your proposed study title, it will help the HREC direct your application through the most appropriate review pathway.

Q1.2 A sample response is provided on the Template HREA form.

Q1.3 A sample response is provided on the Template HREA form.

Q1.4 The response will probably be “Hospital(s), but may be “Clinics” (in addition to or instead of), depending on the source(s) of the data to be used in the proposed study.

Q1.5 This will be the employing organisation of the (Co-ordinating) Principal Investigator.

Q1.6 The response to this will probably be “There is no funding.”

Q1.9 If there is more than one researcher, ensure that one is designated as the contact person and that at least one is authorised to sign on behalf of all the researchers.

If one of the researchers is a student, ensure that the student’s supervisors are also listed as investigators. Most HRECs prefer students NOT to be the (Co-ordinating) Principal Investigator and contact person for a research study.

Q1.9.10 Read the guidance in the form carefully before deciding on each investigator’s position in the proposed study.

Q1.10 The response to this will probably be NO.

Q1.11 The response to this will probably be NO.

Q1.12 If one of the investigators is a postgraduate student, the proposal may have been reviewed by a university committee to ensure its suitability for the program of study in which the student is enrolled.

Q1.13 With the establishment of the National Mutual Acceptance Scheme, it is unlikely that your proposed study has been previously reviewed by another HREC.

Q1.14 If your proposed study specifically involves the collection and analysis of the medical record data of Aboriginal or Torres Strait Islander (ATSI) peoples, or if you wish to collect information on ATSI status and report findings for this group separately from the rest of the study population, you will also need to submit your proposal to the HREC of the Aboriginal Health & Medical Research Council of NSW for review and approval:

 

Similarly, if your proposed study involves the collection and analysis of the medical records of people in the correctional environment of NSW, you will need to submit your proposal to the Justice Health HREC for review and approval:

 

Similarly, if your proposed study involves linkage to state-wide data collections owned or managed by the NSW Ministry of Health or the Cancer Institute NSW, you will need to submit your proposal to the NSW Population & Health Services Research Ethics Committee:

 

Q1.15 If you are proposing to collect data from the medical records held at only one hospital, eg Concord Hospital, then tick NO to this question. If you are proposing to collect data from the medical records held at more than one hospital, even if they are both / all within the Sydney LHD, then tick YES to this question.

Q1.16 The response to this is YES, as you will need to seek research governance authorisation at each site at which you wish to collect data BEFORE commencing the data collection. To do this, you will need to complete the Site Specific Assessment (SSA) form associated with your HREA on the Online Forms website / REGIS website and submit it to the Research Governance Officer at the site at which you wish to collect data. Multiple SSAs can be generated if you are collecting data at more than one site.

Q1.17 It is recommended that you review the information on “What are data? and “Data identifiability” in Chapter 3.2 of the National Statement on Ethical Conduct in Human Research 2007 (updated July 2018).

Non-identifiable data:

If you are collecting non-identifying data, for example anonymised aggregate dataeg via a data dump from electronic records, then it is not necessary to tick any of these boxes. You should provide a list of the data fields which you require as an attachment to your submission. It should include a footer with page numbering (in the format “Page X of Y”), a version number and a date.

Individually identifiable data or re-identifiable data:

If you firstly need to obtain a list of identified patients with a certain condition in order to obtain the non-identifying data dumpaggregate data, then you must tick “Data linkage research”.

Similarly, if you are collecting data from multiple sources, eg from each patient’s eMR and paper MR, then you will need to tick “Data linkage research”, as you will need to collect identifying information in order to link the data from the eMR with the data from the paper MR.

If you need to collect identifiable data, then you will need to provide a Master Code Sheet and a Data Collection Form. Both should include a footer with page numbering (in the format “Page X of Y”), a version number and a date.

Examples of a Master Code Sheet and a Data Collection Form are given in Appendix A and Appendix B.

Q1.18.1 Although it seems counter-intuitive, the correct answer to this question is YES, ie you will be collecting the data from its source in the future.

Q1.19 As a medical record review does not involve any contact with the patients whose data are being used, and it is likely that you will request a waiver of the usual requirement for consent from the patients whose data you wish to use, it is not necessary to tick any of the boxes.

QM3.1 You should consult your colleagues and a biostatistician on an appropriate response to this question.

QM3.2 You should consult your colleagues and a biostatistician on an appropriate response to this question.

QM3.3 Review the publication recommended in the guidance to this question in the HREA. This scenario is unlikely in the context of a review of hospital medical records, even if linkage between electronic and paper records is undertaken. However, depending on the data collected, patients may still be identifiable even without names being used, eg a patient who has undergone a liver transplant and lives in a small rural community. Care will need to be taken when the study results are published to ensure that potentially identifying information about a patient is not divulged.

Q2.2.1 –
Q2.2.4 See suggested responses on the Template HREA form.

Q2.2.6 –
Q2.2.8 See suggested responses on the Template HREA form.

Q2.2.8.1 Use the "SLHD Privacy Compliance Form" to guide your response to this question.

Q2.3.1 The principal risk associated with research involving a review of patient medical records is a breach of privacy.

Q2.3.2 You should discuss the steps that you will take to ensure that there is no breach of patient privacy, eg:

• collecting data in a re-identifiable or non-identifiable format,
• storing the electronic data in REDCap,
• storing paper records in a locked filing cabinet in a locked office on the hospital campus,
• not storing or transferring data between study investigators on devices, such as external hard drives or USB drives, that are easily misplaced, lost, or stolen,
• not identifying small cohorts of patients in publications and presentations.

Q2.4.1 &
Q2.4.2 You should begin by discussing the disease / condition that you are studying, its prevalence in the Australian community, its burden of morbidity and mortality. Explain what you are hoping to learn through your proposed study, how it could improve the situation of those with the disease / condition, how it could affect the Australian healthcare system, etc.

Q2.4.3 As there will be no patient contact, you can answer “Not applicable”.

Q3.1 &
Q3.2 It is recommended that you review the definitions of “personal information’, ‘sensitive information’ and ‘health information’ before completing this question. See the link to the HREA Information Page (https://au.ethicsform.org/Help/Guidance.aspx) in the guidance notes for Q3.1.

You should at least tick “Health information’ for both questions.

Q3.3 Refer to your response to Q1.17 and to the notes above, and ensure that your response to Q3.3 is consistent with that given to Q1.17.

Q3.4 Your answer to this question may be the same as Q3.3 or may indicate that the data to be used in the proposed study have an identifiability that is lower than the data collected, eg re-identifiable data may have been collected, but the key linking identifiers to the data may have been destroyed, so that only non-identifiable data are available to be used in the proposed study.

Q3.5 Review the guidance notes for this question in the HREA form and respond accordingly.

Q3.6 Depending on the location of the medical records to which you wish to have access, you should tick “Medical/health/mental health record” (for hospital records) and/or “Privately held database” (for private practice records).

Q3.7 Review the guidance notes for this question in the HREA form and respond accordingly.

Q3.8 The response to this is YES, as the data were originally collected for clinical purposes.

Q3.8.1 You could argue that medical records in general are a valuable source of information about the success or otherwise of medical treatments. Depending on the aims of the proposed study, you could, for example, discuss the disease / condition under investigation and the fact that there are differing treatment options that need investigating to enhance evidence-based medicine. You could also discuss whether you believe that patients with this disease / condition would expect doctors to try to improve treatments and would likely give their consent to the use of their data if asked.

Q3.9 The response to this will probably be NO. However:

• if you were to make an incidental finding of relevance to a patient’s health, then this should be reported to your supervisor for the appropriate action.

• if you were to find an error in a medical record, then this would require mandatory reporting through the Incident Information Management System (IIMS).

Q3.10 If you are collecting and using non-identifiable data, you should reiterate this here.

If you are collecting and using re-identifiable data, you should discuss your use of a Master Code Sheet on which identifying information will be linked to a Participant Study Number. The Master Code Sheet will be stored securely and separately from the study data. The study data will be collected on a separate Data Collection Form which will record only the patient’s Participant Study Number to protect patient privacy. You should assure the HREC that a person’s identity will not be revealed in any publication arising from the research.

Q3.11 The response to this will probably be NO.

Q3.12 The response to this will be NO, as there is no patient contact involved in the proposed study.

Q3.13 Review the guidance notes for this question in the HREA form and respond accordingly. It may also be appropriate to discuss your proposed response with more senior colleagues. See also the note on Q3.9 above.

Q3.14 Be sure to answer all four aspects of this question. A sample response is given in the template HREA.

NOTE: You are required to securely store your raw data for a minimum of five years after the publication of results in case your findings are challenged. If the data were collected in a re-identifiable format, then this is the format in which they should be stored for this period. At the end of this storage period, it would be acceptable for the Master Code Sheet to be destroyed and the non-identifying data to be stored indefinitely in REDCap, so that it can be made available to other researchers working in the field.

Q3.15 A sample response is given in the template HREA.

Q3.16 The response to this will be NO, as there is no patient contact involved in the proposed study.

Q3.16.2.1 Two sample responses are given in the Template HREA form.

Q3.17 Review the guidance notes for this question in the HREA form and respond accordingly.

Q3.18 Review the guidance notes for this question in the HREA form and respond accordingly.