HLTH 2024 - Research Methods in Health

Introduction – Research problem, background literature review, definitions, gap in knowledge, aims and hypotheses

Introduction should include, with or without subheadings:

• Identification and description of the research problem. Research problem is the practical (i.e., healthcare practice or policy) or conceptual (i.e., ideas) difficulty the research is attempting to examine or solve. The research problem works like a topic.
• Background, what is already known about the research problem including comprehensive summary and critical appraisal of recent, high-quality, relevant literature with citations. Controversies, conflicting evidence and informed opinion should be explored where academic or professional disagreements about theory, evidence or practice are likely for that topic. For some research topics there may be a lack of controversy and no conflicting evidence; for such topics the absence of comparing and contrasting the exSummary and conclusions from existing literature should be provided.
• Clinical population or other population should be readily apparent, and preferably identified explicitly, so readers know whom or what type of person or thing the research is about.
• Gap in knowledge – what is not already known but needs to be known to address or solve the research problem; to improve knowledge policy and practice.
• Research question stated, otherwise sufficiently implied so reader could state the question if a research question is not explicitly stated.
• Definitions of important terms or terminology where these terms are not common knowledge for the target readership.
• Hypotheses or predictions if the research is confirmatory or if specific predictions are justified from background information.
• Aims or objectives for the project: what project is intended to discover or achieve, providing a reason for the research.

Methods – Design

• Methods sections may or may not have subheadings, not essential though preferable. May be called the Method section, no “s”.

• Naming and description of trial design, how the study was set up including that it was randomised trial with control or placebo condition. Includes design information such as whether a parallel (two or more groups receiving the one experimental condition such as treatment or placebo) or crossover trial (each group receives all conditions).

• For clinical trials comparing treatments with other treatments or control or placebo conditions there are additional design descriptions for clinical trials:

• Equivalence – tests whether or not the treatment gives the same effect as alternatives such as control or placebo.

• Superiority – tests whether or not treatment gives a better effect than alternatives such as control or placebo.

• Non-inferiority – tests whether or not treatment gives an effect no worse than alternatives such as control or placebo

Methods – Setting and study date range

Geographical setting, such a city, district, region, state and country.

• Type of facility or situation where treatments or assessments or other interaction with participants occurred, such as hospital outpatient consulting room, inpatient ward, private clinic, participants’ homes, university campus, residential aged care facility (i.e., nursing home). The name of the facility does not have to be supplied and may reasonably be withheld to maintain privacy.

• Starting and finishing dates especially the year and preferably the month for operational stages (not planning or analysis) study. Start and finish dates are optional, not essential, for good reporting. Ideally, the year in which the study was conducted and data collection should be stated.

Methods – Sample characteristics, selection criteria and relevant measurements or assessments

• Detailed description of the target (e.g., clinical) population as planned. Sample description must be in enough detail for readers to assess how well the sample resembled readers’ own client population. If the sample was a census or intended as a census, that should be apparent.

• Planned demographic descriptors (i.e., general characteristics used for eligibility and selection but may not be clinically relevant).

• Relevant clinical characteristics of the sample.

• Eligibility criteria: inclusion and exclusion listed.

• Any measurements or assessment procedures used to select participants or determine eligibility, including types of people who the assessing, whether research team or other staff. 

Methods – Planned, recruited and analysed group sample sizes – see also Participant Flow, Losses and Attrition, p. 13.

• Sample size at the study outset, when participants are selected and assigned to groups, and after data collection (OK if in Results) should be reported.

• Planned or recruited sample size, and actual sample size analysed after attrition if attrition occurred.

• Numbers of participants lost to the study (i.e., attrition) must be apparent from reading either Methods and Results, or both.

• Reasons for attrition if any occurred – analysed sample size will be smaller than the initially recruited sample size.

Methods – Sampling method

The planned method for recruiting the sample, stated using formal terms such as:

• Random sample – rare in health research because only volunteers

• Convenience sample – such as volunteers self-selecting in response to an advertisement.

• Purposive, judgemental or theoretical sampling – choose exactly the right sort of person from already known characteristics, often used in qualitative research.

• Direct invitation – asking people, inviting people to participate; people chosen from list of likely eligible and willing people.

• Consecutive cohort sampling – selecting people in the same order as they present, such as for medical consultations.

Methods – Sampling procedures, recruitment and retention

How (i.e., procedures, what is actually done and by what member of research team or staff) participants were recruited and enlisted.

• How potential participants were initially found, if they were.

• How initial contact was made.

• How participants were encouraged to join the study.

• Procedures for increasing retention; reducing attrition.

Rewards or incentives for enlisting or continuing participation, retention. Retention may not be an issue for very brief projects. Enough detail to enable replication or identify possible validity threats attributable to recruitment methods should be provided.

Methods – Randomisation

• How randomisation was done, including how the sequence was devised (e.g., computer-generated or sealed envelopes?)

• How randomisation sequence was concealed to protect internal validity. The sequence should be concealed from research staff doing the allocating, so participants are allocated to groups unpredictably.

Methods – Sample size calculations and power analysis

How the sample size was decided.

• Power analysis is expected in this section. Properly reported power analysis should include:

• Minimum effect size considered acceptable, a clinical judgement, not a statistical fact.

• Decision rule for statistical significance, almost always p < .05.

• The desired power, usually at least an 80% chance of a statistically significant result. ▪ The statistical test used.

Methods – Blinding

Whether and how blinding was conducted for these people in the project:

• Staff who allocate participants to treatment, control, placebo, alternative treatment, usual care or whatever groups in the design.

• Patients getting allocated to groups.

• Treating clinicians, those administering treatments or pseudo treatments to patients; anyone working with the patients clinically.

Staff measuring outcomes or conducting assessments or other staff interacting with participants, assessors.

Methods – Interventions; procedures for interventions

• Interventions (i.e., procedures, what actually done for the interventions) for all groups (not only the main intervention) described in enough detail for replication and reader evaluation of relevance, including how control or placebo or no-treatment groups were managed.

• Whether the clinicians administering the treatments and the staff interacting in other ways with the participants were the principal or associate researchers or were research assistants employed directly by the project (i.e., part of the research team), or were personnel employed separately from the project team should be stated or apparent.

Methods – Equipment, apparatus

• Apparatus or equipment used to implement treatments, described in enough detail so other clinicians can evaluate for appropriateness and use the same or compare for similarity with own equipment.

• Equipment used to administer treatments, pseudo treatments, placebo or usual care.

• Equipment that researchers and participants use to conduct tasks, where the selection of that equipment could influence results, including measurement reliability or validity, or overall internal or external validity, generalisability, applicability, or research ethics.

• Equipment used to measure outcomes or other assessment procedures including for selecting the sample where equipment was used. Includes technical equipment such as measurement tools, questionnaires, surveys, interview questions and rating scales. Standardised tests should be identified by name and citation, and preferably include published evidence of reliability and validity.

Methods – Outcomes measurement and methods

• Outcome measures defined and described in enough detail to enable replication and for reader to assess relevance.

• Procedures used obtain measurements of outcomes, in enough detail to enable replication.

• Whether the clinicians conducting assessments were the researchers, other project staff or persons employed independently from the project should be stated or apparent.

• Primary outcome (usually only one but can be a few primaries), the main outcome measure of interest to the project.

• Secondary outcome are additional outcomes of interest, relevant though less important to study aims than primary outcomes.

• Any changes to choice of outcomes or measurement procedures should be document.

Methods – Statistical methods

Statistical methods to analyse the date to describe and compare groups for primary and secondary outcomes, plus any extra of follow-up analysis beyond what answers the main research questions (PICO). This section should identify descriptive statistics and statistical significance (inferential) tests used. Statistical procedures should be listed. Description of statistical methods should match statistics used in the Results.

Whether intention to treat or per protocol analysis was conducted if there was a lot of attrition. May not be necessary to mention if there was little or no attrition.

Methods – Ethics

Ethical topics relevant to the study and how they were managed, including participant information, consent, anonymity, ethically acceptable compensation, reward or inducement, confidentiality, harm identification and risk minimisation; official approval.

Methods or Results – Participant flow, losses and attrition – see also Planned, Recruited and Analysed Group Sample Sizes, p. 7

• Flow diagram with numbers of participants who were randomly assigned, received intended treatment, and were analysed for outcomes.

• The flow diagram should show losses and attrition, if any, and whether and how any, if any, participants changed groups.

• Location of flow diagram in the article, either in Methods or Results, has no implications for reporting quality or healthcare practice.

• Reasons for losses to the sample (recruited participants withdrawing from the study) before and after randomisation should be given. Reasons for losses may be provided in Results or Methods; either are acceptable reporting.