Randomized Trials Using Provincial Health Numbers for Group Assignment

Area of Science:

• Health Services Research
• Clinical Trials Methodology
• Health Informatics

Background:

• Provincial health card numbers in Ontario, Canada, are unique, lifelong identifiers assigned randomly.
• These numbers are 10-digit strings with a version code, subject to proprietary alterations before assignment.
• The random generation and unique nature of health card numbers present an opportunity for trial randomization.

Purpose of the Study:

• To demonstrate the feasibility of using Ontario health card numbers for individual-level randomization in large pragmatic trials.
• To analyze the distribution of health card digits for randomization purposes.
• To discuss the practical and methodological considerations of this randomization approach.

Main Methods:

• Analysis of health card number assignment and digit distribution in a trial setting.
• Development of a randomization method utilizing specific digits of the health card number (e.g., the tenth digit).
• Examination of the implications for trial integrity and stratification.

Main Results:

• Health card numbers can be effectively used for individual-level randomization in large pragmatic trials.
• The tenth digit of the Ontario health card number was utilized for randomization in a trial of dialysate bicarbonate concentrations.
• The approach offers practical advantages, including streamlining processes and reducing trial costs.

Conclusions:

• Using provincial health card numbers for pragmatic randomized trials is a viable strategy.
• The benefits of this method, such as cost reduction and process streamlining, must be weighed against potential impacts on randomization integrity and stratification.
• Careful consideration on a trial-by-trial basis is essential before implementing this approach.