Question

We conducted a 2-year retrospective cohort study among all elderly members of a staff-model managed care organization who had a baseline diagnosis of chronic lung disease. The study outcomes were assessed over 2 years, from November 15, 1993, through November 14, 1995, and included hospitalizations for pneumonia and influenza. Of 1898 subjects, 1280 (67%) had received pneumococcal vaccination. This included 843 (44%) who were vaccinated prior to November 15, 1993 and an additional 437 (23%) vaccinated after that date. During the follow-up period there were 174 hospitalizations for pneumonia and influenza. The observed cumulative incidence was 138 per 1000 in the 618 unvaccinated persons and 70 per 1000 in the 1280 persons following their receipt of vaccination. This represents a 49% reduction in hospitalization for pneumonia and influenza. Even if there were neither misclassification nor confounding in this study, the estimate of benefit associated with pneumococcal vaccination in elderly patients with chronic lung disease must be biased. a. Why? In which direction? b. How could the analysis be conducted to remove this source of bias?

Answer 1

a.) Estimate of pneumococcal vaccination in elderly patients with chronic lung disease is biased. Especially in the direction of healthy vaccinated adults. Because, more than 80% of healthy adults who received vaccination develop antibodies against the serotypes contained in the vaccine. This immune response occurs within 2 to 3 weeks after vaccination. Also, older adults and people with chronic illnesses or immunodeficiency may not respond as well. Elevated antibody levels persist for atleast 5 years in healthy adults, but decline more quickly in elderly and people with chronic illness.

b) In a retrospective cohort study, both exposure(vaccination) and outcome(disease) have occured when the study begins. Only exposure is taken into consideration and the develeopment of same disease is looked for in exposed(vaccinated) and unexposed(unvaccinated).

Incidence in vaccinated = 70

Incidence in unvaccinated = 138

So, Relative risk = 0.51

A case control study can be taken to prevent the bias. Or, same age group people can be taken as the vaccination effect differs in elderly, adults and chidren.