Understanding Berksonian Bias

Berksonian bias is a specific type of selection bias. It arises because the probability of being selected for a study is associated with both the exposure and the disease being investigated, specifically within a hospital setting.

Berksonian Bias in Study Types

This bias primarily occurs in studies conducted on patients admitted to hospitals (hospital based studies). The reasoning is as follows:

• Hospitals primarily admit individuals who are diseased.
• The frequency of certain exposures (like smoking or specific dietary habits) might differ between those admitted to the hospital versus the general population.
• For example, if a specific exposure is more common in individuals who are likely to become severely ill and require hospitalization, the study population (hospital patients) will have a higher proportion of exposed individuals compared to the general population, even if the exposure doesn't directly cause the disease.
• This distortion in the exposure-disease relationship happens because the sampling frame (hospital admissions) is inherently biased towards certain conditions and potentially certain exposures linked to those conditions.

Therefore, Berksonian bias specifically affects the representativeness of samples drawn from hospital populations, impacting the generalizability of findings related to disease etiology.

Why Not Other Study Types?

• Community based studies aim to sample from the general population, reducing the hospital admission bias.
• Laboratory based studies focus on biological mechanisms and typically don't involve patient selection bias in the same way observational studies do.
• Natural exposure studies, depending on their design, may occur in various settings but the term "natural exposure" itself doesn't inherently imply the specific selection mechanism seen in hospitals.

The core issue is the non-random selection of diseased individuals needing hospital care.