Evaluating ADHD screening tools: A comparative analysis of accuracy, cost, and complexity

Debate over the rising prevalence of ADHD diagnoses underscores the need for accurate screening. This study provides clinicians a comparative inventory of tools, evaluating balanced accuracy, item count, cost, reporter type, and predictive values at realistic base rates.

Tools were identified via systematic searches of MEDLINE, PsycINFO, and EMBASE (Jan 2021-July 2025), extending prior reviews. Instruments for children, adolescents, and adults were characterized by reporter type. Balanced accuracy was modeled using mixed-effects beta regression with a random intercept for tool and fixed effects for reporter type, cost, and item count; minimum age was included in sensitivity analyses. Positive/negative predictive values (PPV/NPV) were also examined.

Seventy-four assessments covering 40 reporter-based screening tools were analyzed (mean balanced accuracy 76 %, SD = 0.09). Self-report tools outperformed teacher reports; differences were not significant after adjusting for age. Cost showed no association with balanced accuracy; item count showed a small, non-significant positive trend (p = 0.06). At 10 % prevalence, PPV was modest (0.10-0.30) and NPV was high (0.95-0.99).

ADHD screening tools showed fair to good accuracy. Brief tools can perform well, with no observed advantage for longer or paid tools. Reporter-type effects were partially explained by age. Given low PPV but high NPV at base rates, positive screens warrant full assessment, while negative screens could reliably rule out most ADHD cases. Our findings suggest that the selection of ADHD screening tools should be guided by population, content, and context rather than by length or cost.