Research studies gather data, analyze it, and draw conclusions. The possible topics of study are almost limitless!
Note: Any research on humans requires strict privacy safeguards and Ethics Board review.
Research studies vary considerably in their size and complexity. Some research topics require large sample populations to show statistical significance. Prospective trials may require a lot of time to recruit subjects, trial the intervention, then gather the data and analyze the results. If the ethics board deems the project to be more than “Minimal Risk” much more time may be required for design, approval, and implementation. These and other factors will put some project ideas out of your reach, but, on the other hand, if you start early (eg. in R1 year), collaborate with others, and keep the focus on a small objective, many of the following research study types are feasible as UBC Resident Scholar Projects.
1. Primary data source (using data you have collected)
This type of data is good for eliciting a subject’s experience, feelings, and expectations about an issue or intervention is usually taped or recorded then analyzed using qualitative methods like grounded theory. Some examples include:
- self-reflection (Eg. experience as an IMG resident in Canada)
- journaling (Eg. experience working with inner city poor)
- individual or group interviews (Eg. determining the factors affecting a patient’s choice to go to sexual health clinic rather than their family doctor)
- video or live observation (Eg. determining if a waiting room toy box is utilized by sick children)
- creative arts as data (Eg. using photos, videos, art or personal writings to gather information about subjects and themes for analysis)
This type of data is what we normally elicit when we have a PICO question. It involves measurement of objects, environment, human clinical parameters, surveys, etc. There may or may not be an intervention with a therapy, a test, a natural event, and it may be controlled or not, blinded or not. The study population size is determined in advance in order to achieve statistical significance. Some examples include:
- survey – this is a cross-sectional analysis of opinion or experience. It can be repeated after an intervention to make comparisons. (Eg: patient satisfaction with care)
- cohort – this is where you measure a population repeatedly over time, perhaps before and after an intervention (Eg. determining the effectiveness of a new weight-reduction program)
- trial – this is where you trial a therapy or validate a tool (link to Validation) with a control group for comparison, preferably blinded. (Eg Randomized Controlled Trial to determine if a new wart treatment is better than placebo, or RCT to validate a streptococcal throat predictor tool)
2. Secondary data source (using data collected by someone else)
Personal or Private Data
(Includes hospital or office medical records, Pharmacare or Cancer Screening Program data, etc. Access to and use of this data often requires consent and/or Ethics Committee approval)
- case report (Eg. unusual case presentations in family practice)
- audits and quality improvement (Eg. Before and after chart audits to show improvement in diabetic management with a new office EMR tool)
- Case Control Research Study (retrospective) (Eg. Chart review to show whether patients given the QuitNow smoking cessation package were more likely to quit)
- Secondary analysis of results (Eg. Pharmacare prescription data for patterns of use in )
- Correlation between databases (Eg. determine if MVA deaths are more common in regions of high narcotic prescription use)
Anonymous or Public Data
(Includes vital statistics, census, government and health authority financial reports, National Physician Survey results etc. This data is public and sometimes freely available on the internet or by permission through other means. Use of this data does not require Ethics Committee approval.)
- Secondary analysis of results (Eg. Analysis of NPS results to determine if confidence in doing procedures at the end of residency is correlated with future rural practice location)
- Correlation between databases (Eg. determine if suicide rates are higher in areas of low socio-economic status)
- Meta-analysis including systematic review. (Eg. determine if the weight of research evidence favors e-cigarette use for smoking cessation)
To plan a research study, you need a protocol. A template protocol that would fit most research studies is available here. It is important that you write up your draft protocol and send it to your Site Research Faculty as soon as possible.