Understanding the ADHD Knowledge–Attitude Gap Among Physicians and Senior Medical Students in Jordan [Letter]

Dear editor

We read with great interest the article by Al-Ghawanmeh et al entitled “Knowledge and Attitudes Toward Attention-Deficit/Hyperactivity Disorder Among Physicians and Senior Medical Students in Jordan”.¹ The authors should be congratulated for addressing an important issue in medical education and neurodevelopmental health and for including both senior medical students and practicing physicians in their survey. Their effort to examine knowledge and professional attitudes together is particularly relevant because clinical competence in ADHD extends beyond factual knowledge alone. We would, however, like to raise several methodological and statistical points that may help readers interpret the findings more cautiously.

A first concern relates to the questionnaire. The manuscript states that the instrument was adapted from previously validated questionnaires and modified according to DSM-5 criteria. Although content validity and Cronbach’s alpha coefficients were reported, key psychometric details were not provided, including the number and expertise of the reviewers involved in content validation, item-level validity indices, structural validity analyses, and test–retest reliability. COSMIN guidance recommends that modified instruments undergo renewed psychometric evaluation because changes in wording or diagnostic criteria can affect both validity and reliability.^2,3 More complete reporting of the validation process would therefore strengthen confidence in the measurement of ADHD-related knowledge and attitudes.

The scoring procedure also deserves clarification. Correct responses were scored +1, incorrect responses −1, and “not sure” responses 0. Negative marking is uncommon in knowledge assessment and can influence score distributions and participant ranking differently from conventional dichotomous scoring.⁴ In addition, the manuscript refers to predefined categories for poor, fair, and good knowledge and for positive versus negative attitudes, but the actual cutoff values were not reported. Providing these thresholds is essential for reproducibility and for meaningful interpretation of the reported prevalence estimates.³

The statistical analysis raises a more substantive issue. Variables were entered into the multivariable ordinal logistic regression models only if they were significant in bivariate analyses. This data-driven approach can omit clinically important confounders and may produce unstable adjusted estimates.⁵ More importantly, the authors reported that the proportional odds assumption for the student model was violated (test of parallel lines p<0.001). Because ordinal logistic regression relies on this assumption, retention of the cumulative logit model despite a significant test is difficult to justify statistically. Alternative approaches such as multinomial logistic regression or partial proportional odds models would have been more appropriate.⁶ The conclusion that completion of a pediatric rotation independently predicted higher knowledge therefore rests on a model whose main assumption was not satisfied.

Interpretation of the physician subgroup analysis also requires caution. Some categories were very small, including psychiatry and several specialist groups. Sparse subgroup data can lead to unstable regression estimates and wide confidence intervals.⁷ This is particularly relevant to the reported association between having children and higher knowledge, which reached statistical significance (p=0.049) but was accompanied by a wide confidence interval (95% CI 1.003–10.363), indicating substantial imprecision. A more restrained interpretation would better reflect the uncertainty around this estimate.

We also noted that numerous bivariate comparisons were performed without adjustment for multiple testing. Repeated hypothesis testing increases the likelihood of false-positive findings, especially in exploratory subgroup analyses.⁸ Clarifying whether the analyses were exploratory or reporting adjusted p-values would improve statistical transparency.

Another point concerns the presentation of the results. The proportions of participants with good knowledge and positive attitudes were reported without confidence intervals. STROBE recommends reporting measures of precision for principal estimates so that readers can judge sampling variability and estimate stability.⁹ Inclusion of 95% confidence intervals would strengthen the presentation of the prevalence findings.

Finally, the discussion places considerable emphasis on the educational value of pediatric rotations. Given the violation of the proportional odds assumption in the student regression model, this educational interpretation should be presented more cautiously. The observed association is useful for generating hypotheses, but it should not be interpreted as evidence that pediatric rotations improve ADHD knowledge. Longitudinal or interventional studies would be required to evaluate such an educational effect directly.^6,9

Despite these concerns, the study provides useful preliminary data on ADHD-related knowledge and attitudes among physicians and senior medical students in Jordan. Additional psychometric reporting, clearer score construction, regression models that satisfy statistical assumptions, and fuller presentation of precision estimates would further strengthen future research in this important area.

Funding Statement

The authors received no financial support for the preparation of this Letter to the Editor.

Data Sharing Statement

No new data were generated or analyzed for this correspondence.

Ethical Approval

Ethical approval was not required for this Letter to the Editor because it is based exclusively on analysis of a previously published article.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no conflicts of interest in this communication.