Loading Tree…

DQI-2004

Definition

The proportion of all participants who only partially complete the study and prematurely abandon it.

Explanation

Drop-out has been defined as an indicator because of its high importance within longitudinal study designs to address individuals who prematurely abandon a study.

Contrary to approaches by AAPOR, full and partial responses are counted alike at this level because partial responses are dealt with at the segment or item level.

Example

During the recruitment of individuals for a 10-year follow-up examination in a cohort study the following distributions emerge for 600 participants:

  • I, P: Complete or partially participate in the follow-up examination (N=400)

  • R: Refusal (N=10)

  • NC: Non-contacts (e.g. target participant with known eligibility unavailable) (N=20)

  • O: Other eligible no-responder (e.g. is unavailable at projected examination dates) (N=50)

Unknown eligibility, no examination

  • NE: Not eligible (e.g. deceased) (N=120)

In total, 400 out of 480 eligible subjects participate, leading to a drop-out rate of 18%

In contrast, a crude missing value indicator at the level longitudinal missingness would have indicated a proportion of 33% missing subjects without follow-up based on 200 out of 600 individuals who did not participate in the follow up examination.

Guidance

Drop out-rates should be computed whenever there is a full available coding of missing data values in a longitudinal study.

In case of uncertainties about eligibility, AAPOR suggests to assume that a certain proportion of all observational units with unknown eligibility is not-eligible and should be removed from the denominator.

Drop out-rates are always equal or lower compared to corresponding missing value percentages.

Interpretation

The higher the drop-out rate, the lower the data quality.

Literature