Patient Health Questionnaire-9 Item Pairing Predictiveness for Prescreening Depressive Symptomatology: Machine Learning Analysis
Nenhuma Miniatura disponível
Citações na Scopus
0
Tipo de produção
article
Data de publicação
2023
Título da Revista
ISSN da Revista
Título do Volume
Editora
JMIR PUBLICATIONS, INC
Autores
GLAVIN, Darragh
GRUA, Eoin Martino
SANTOS, Edinilza Ribeiro dos
WONG, Gloria H. Y.
HOLLINGWORTH, William
PETERS, Tim J.
ARAYA, Ricardo
VEN, Pepijn Van de
Citação
JMIR MENTAL HEALTH, v.10, article ID e48444, 19p, 2023
Resumo
Background: Anhedonia and depressed mood are considered the cardinal symptoms of major depressive disorder. These are the first 2 items of the Patient Health Questionnaire (PHQ)-9 and comprise the ultrabrief PHQ-2 used for prescreening depressive symptomatology. The prescreening performance of alternative PHQ-9 item pairings is rarely compared with that of the PHQ-2.Objective: This study aims to use machine learning (ML) with the PHQ-9 items to identify and validate the most predictive 2-item depressive symptomatology ultrabrief questionnaire and to test the generalizability of the best pairings found on the primary data set, with 6 external data sets from different populations to validate their use as prescreening instruments.Methods: All 36 possible PHQ-9 item pairings (each yielding scores of 0-6) were investigated using ML-based methods with logistic regression models. Their performances were evaluated based on the classification of depressive symptomatology, defined as PHQ-9 scores >= 10. This gave each pairing an equal opportunity and avoided any bias in item pairing selection.Results: The ML-based PHQ-9 items 2 and 4 (phq2&4), the depressed mood and low-energy item pairing, and PHQ-9 items 2 and 8 (phq2&8), the depressed mood and psychomotor retardation or agitation item pairing, were found to be the best on the primary data set training split. They generalized well on the primary data set test split with area under the curves (AUCs) of 0.954 and 0.946, respectively, compared with an AUC of 0.942 for the PHQ-2. The phq2&4 had a higher AUC than the PHQ-2 on all 6 external data sets, and the phq2&8 had a higher AUC than the PHQ-2 on 3 data sets. The phq2&4 had the highest Youden index (an unweighted average of sensitivity and specificity) on 2 external data sets, and the phq2&8 had the highest Youden index on another 2. The PHQ-2 >= 2 cutoff also had the highest Youden index on 2 external data sets, joint highest with the phq2&4 on 1, but its performance fluctuated the most. The PHQ-2 >= 3 cutoff had the highest Youden index on 1 external data set. The sensitivity and specificity achieved by the phq2&4 and phq2&8 were more evenly balanced than the PHQ-2 >= 2 and >= 3 cutoffs.Conclusions: The PHQ-2 did not prove to be a more effective prescreening instrument when compared with other PHQ-9 item pairings. Evaluating all item pairings showed that, compared with alternative partner items, the anhedonia item underperformed alongside the depressed mood item. This suggests that the inclusion of anhedonia as a core symptom of depression and its presence in ultrabrief questionnaires may be incompatible with the empirical evidence. The use of the PHQ-2 to prescreen for depressive symptomatology could result in a greater number of misclassifications than alternative item pairings.
Palavras-chave
Patient Health Questionnaire-2, PHQ-2, Patient Health Questionnaire-9, PHQ-9 items, depressive symptomatology, ultrabrief questionnaires, prescreening, machine learning, cardinal symptoms, low energy, psychomotor dysfunction, depressed mood
Referências
- American Psychiatric Association, 2013, Diagnostic and statistical manual of mental disorders, 5th edition (DSM-5), Vfifth, DOI 10.1176/APPI.BOOKS.9780890425596
- Arrieta J, 2017, J CLIN PSYCHOL, V73, P1076, DOI 10.1002/jclp.22390
- BURNAM MA, 1988, MED CARE, V26, P775, DOI 10.1097/00005650-198808000-00004
- dos Santos ER, 2016, PLOS ONE, V11, DOI 10.1371/journal.pone.0150046
- El-Koka A, 2013, INT CONF ADV COMMUN, P13
- Feng LX, 2022, J LEARN DISABIL-US, V55, P427, DOI 10.1177/00222194211047631
- Glavin Darragh, 2022, Procedia Computer Science, P101, DOI 10.1016/j.procs.2022.09.089
- Henkel V, 2004, EUR ARCH PSY CLIN N, V254, P215, DOI 10.1007/s00406-004-0476-3
- Johnson JG, 2002, J ADOLESCENT HEALTH, V30, P196, DOI 10.1016/S1054-139X(01)00333-0
- Kennedy Sidney H, 2008, Dialogues Clin Neurosci, V10, P271
- Kim S, 2021, INT J ENV RES PUB HE, V18, DOI 10.3390/ijerph18073339
- Kroenke K, 2003, MED CARE, V41, P1284, DOI 10.1097/01.MLR.0000093487.78664.3C
- Kroenke K, 2001, J GEN INTERN MED, V16, P606, DOI 10.1046/j.1525-1497.2001.016009606.x
- Levis, 2019, BMJ-BRIT MED J, V365, DOI 10.1136/bmj.l1781
- Levis B, 2020, JAMA-J AM MED ASSOC, V323, P2290, DOI 10.1001/jama.2020.6504
- Levis B, 2017, AM J EPIDEMIOL, V185, P954, DOI 10.1093/aje/kww191
- Liu TY, 2022, TRIALS, V23, DOI 10.1186/s13063-022-06122-1
- Lopes CD, 2022, CAD SAUDE PUBLICA, V38, DOI [10.1590/0102-311X00123421, 10.1590/0102-311x00123421]
- Lopez TJ, 2018, Dataset
- Löwe B, 2005, J PSYCHOSOM RES, V58, P163, DOI 10.1016/j.jpsychores.2004.09.006
- Lux V, 2010, PSYCHOL MED, V40, P1679, DOI 10.1017/S0033291709992157
- Manea L, 2016, J AFFECT DISORDERS, V203, P382, DOI 10.1016/j.jad.2016.06.003
- Mitchell AJ, 2009, PSYCHOL MED, V39, P1107, DOI 10.1017/S0033291708004674
- Mitchell AJ, 2016, BJPSYCH OPEN, V2, P127, DOI 10.1192/bjpo.bp.115.001685
- Paykel Eugene S, 2008, Dialogues Clin Neurosci, V10, P279
- Rosenström T, 2015, INT J METH PSYCH RES, V24, P213, DOI 10.1002/mpr.1478
- Scazufca M, 2022, LANCET HEALTH LONGEV, V3, pE690, DOI 10.1016/S2666-7568(22)00194-5
- Scazufca M, 2020, TRIALS, V21, DOI 10.1186/s13063-020-04826-w
- Scazufca M, 2016, PLOS ONE, V11, DOI 10.1371/journal.pone.0157719
- Smits N, 2010, BMC MED RES METHODOL, V10, DOI 10.1186/1471-2288-10-89
- SPITZER RL, 1994, JAMA-J AM MED ASSOC, V272, P1749, DOI 10.1001/jama.272.22.1749
- Van de Ven P, 2019, J MED INTERNET RES, V21, DOI 10.2196/11346