Selecting a cutoff point for a developmental screening test based on overall diagnostic indices and total expected utilities of professional preferences.

Background/purpose: A cutoff point in a test with sounded validity and professional preferences can help to make an accurate clinical decision. This study aimed to determine a cutoff point between two strategies for a developmental screening checklist (referred to as Taipei II). Cutoff point A was set as one or more item failed and cutoff point B was set as two or more items failed or one or more marked item failed.

Methods: This study was based on the total expected utilities of professional preferences and overall diagnostic indices. A self-administered questionnaire was developed to collect the estimated utility from professionals involved in early childhood interventions (n = 81) regarding four screening outcomes (probabilities of true positive, false positive, true negative, or false negative) and costs. The total expected utilities were calculated from the probabilities of four screening outcomes and utility values.

Results: The diagnostic odds ratio was higher for strategy B (695 and 209, respectively) than that of strategy A (184 and 150, respectively) when using the Taipei II on children under 3 years of age and age 3 and over. Strategy B also had a higher median total expected utilities score than strategy A (0.78 vs. 0.72 for children < 3 and 0.76 vs. 0.67 for children > or = 3).

Conclusion: If only one cutoff point can be chosen, the authors suggest that clinicians should choose cutoff point B when using the Taipei II for screening. However, two cutoff points of Taipei II, a combination of strategy A and B, can also be used clinically.