Observational Studies: Introduction and Clinical Application

Speaker: Fiona Blythe, Director of the Master of Clinical Epidemiology program in the School of Public Health.
Scope: This is the first of a series of four lectures focused on observational study designs.
Learning Outcomes: - Describe the primary differences between observational and experimental study designs. - Define specific clinical questions where observational studies are considered useful. - Identify common types of observational studies encountered in clinical research. - Understand the essential role of critical appraisal regarding observational studies.

Foundational Knowledge: Previous studies have covered the broad architecture of study types, with a recent focus on experimental designs, specifically Randomised Controlled Trials (RCTs).
Core Distinction: The fundamental difference is that in observational study designs, there is no element of randomisation or purposeful allocation of participants to different groups.
Hierarchy of Evidence: Because they lack randomisation, observational studies provide a lower level of evidence than RCTs when the research question concerns the effect of an intervention.

Ethical Constraints: Used for studying harmful exposures where it would be unethical to randomise participants (e.g., exposing individuals to known toxins or risks).
Practical Constraints: Used when randomisation is not practical, such as when studying variables that cannot be assigned, such as a person's age group or ethnic background.
Duration and Frequency: Observational designs are effective for measuring outcomes that are rare or take a very long period of time to develop.
Specific Clinical Question Types: - Risk Factors: Preferred for identifying factors that contribute to the development of a disease. - Prognostic Factors: Preferred for answering questions about the likely course of a disease. - Descriptive Questions: Useful for determining prevalence (how common a condition is at a point in time) or incidence (rate of new cases) in a given population.
Strategic Selection: Researchers must understand the specific type of clinical question being asked to choose the optimal study design.

Current and upcoming observational study topics include: - Cohort Studies (dedicated online lecture). - Case Control Studies (dedicated online lecture). - Other Observational Designs: Covers the remainder of commonly used study types.

Historical Context: Traditionally, observational studies were primary research studies involving direct participant recruitment and data collection by the researcher.
Modern Shift: Increasing use of data collected for purposes other than research, known as Routinely Collected Health Data (RCD).
Primary Examples of RCD: - Disease registries. - Health service administrative data sets.
Local Data Sources (New South Wales): - New South Wales (NSW) admitted patient data collection. - NSW Register of Births, Deaths and Marriages.
Hybrid Methodologies: RCD can be used as standalone data sets or integrated into primary observational studies through data linkage methods.

Research Question: What are the factors associated with long-term opioid use for chronic non-cancer pain in the New South Wales (NSW) adult community?
Data Set Construction (Three Sources): 1. Primary Cohort Study: Included more than $260,000$ individuals aged $45$ and over living in NSW during a defined period (the 45 and Up Study). This is noted as the largest cohort study in the Southern Hemisphere. 2. Pharmaceutical Benefits Scheme (PBS) Data: Participant data was linked (with permission) to these records to track all prescriptions over the study period. 3. NSW Registry of Births, Deaths and Marriages: Linked to determine if study dropouts were due to participant mortality.
Principal Findings: - Long-term opioid users were more likely to have specific healthcare access characteristics, including low income status and living outside major cities. - Health characteristics associated with long-term use included lower self-reported physical functioning, higher levels of psychological distress, and osteoarthritis. - These findings were compared against participants who used opioids only episodically or for short durations.

Rationale: The "explosion of information" in research makes it necessary to efficiently judge which studies are reliable and useful.
Tools and Guides: Critical appraisal tools focus on the elements of study design and execution specifically linked to the risks of common forms of bias.
Application: These tools allow for efficient judgments when assessing research to answer specific clinical questions, such as those used in "rapid evidence into practice" group activity sessions.

Historical Issues: A historical lack of consistency in reporting made it difficult to find basic information on study design and conduct in peer-reviewed journals.
Current Progress: Reporting guidelines have been developed for various study types and are often required by journals for publication.
Function: While not critical appraisal tools themselves, reporting standards help readers navigate papers during the appraisal process.
The Equator Network: A centralized website that compiles reporting guidelines for all major study types, helping to strengthen the evidence base for patient care.

Summary of Lecture 1: - Key difference between RCTs and observational studies is the absence of randomisation. - Defined the utility of observational studies for risk, prognosis, and prevalence. - Introduced the framework for critical appraisal.
Upcoming Content: - Features, strengths, and limitations of specific observational designs. - Common types of bias relevant to each design. - Strategies for limiting or mitigating bias.