Headache diagnosis in children and adolescents: validation of the Italian version of the HARDSHIP questionnaire

Abstract

Background

Headache disorders are common in children and adolescents, significantly affecting quality of life and academic performance. Despite their high prevalence, they remain underdiagnosed and undertreated. The Child and Adolescents HARDSHIP questionnaire, developed by the Global Campaign Against Headache, has been widely used in adult epidemiological studies but lacks validation in the Italian pediatric population.

Methods

The study was conducted within the Epidemiologic Registry of Child and Adolescents Headaches - Registro Epidemiologico delle Cefalee in età evolutiva (REPICEF), a population-based registry on pediatric headache epidemiology in L’Aquila, Italy. The Italian version of the HARDSHIP questionnaire was adapted using the TRAPD method, ensuring linguistic and conceptual accuracy. Validation was performed by comparing questionnaire-based diagnoses with clinical diagnoses attributed by expert neurologists assumed as gold standard. We computed the sensitivity, specificity, positive and negative predictive values.

Results

Out of 858 screened children and adolescents, the first 535 (62.4%) were selected for the validation questionnaire. Based on the HARDSHIP questionnaire, the most common diagnoses were probable Migraine (pMig) in 175 (20.7%), probable Tension-Type Headache (pTTH) in 144 (16.7%), Undifferentiated Headache (UDH) in 105 (12.2%), Tension-Type Headache (TTH) in 86 (10.0%), migraine in 68 (7.9%) children and adolescents. The agreement between questionnaire-based and clinical diagnoses was for migraine k = 0.432, for TTH k = 0.327, for pMOH k = 0.214.

Conclusions

The validated Italian version of the HARDSHIP questionnaire provides a useful instrument for epidemiological research and for the improvement in diagnosis of pediatric headache disorders.

Supplementary Information

The online version contains supplementary material available at 10.1186/s10194-025-02113-0.

Background

According to the Global Burden of Disease (GBD) study 2021, headache disorders rank 15th among the causes of disability-adjusted life years (DALYs) worldwide and 9^th among children aged 5–14 years [1]. Epidemiological studies performed in various countries have consistently reported a high prevalence of headache disorders in individuals aged 6–17 years. A meta-analysis estimated the overall prevalence of primary headaches in children at 62% but underscored the need for further accurate epidemiological research in pediatric populations [2].

Despite their substantial impact on quality of life [3–5] and school performance [6, 7], headache disorders remain underdiagnosed and undertreated among children and adolescents [2, 8]. Headache type misdiagnosis is an additional substantial problem in children and adolescents due to atypical features compared to adult forms and the lack of standardized diagnostic criteria in those age groups [9]. The accuracy of self-reported headache symptoms among children and adolescents is limited due to developmental constraints in symptom recognition, recall, and communication skills, posing additional challenges for diagnosis and treatment. Those challenges limit epidemiological research on pediatric headache disorders, also restricting our understanding of its burden and progression over time. To address this gap, the Global Campaign Against Headache, promoted by the Lifting the Burden consortium, developed the Child and Adolescents Headache-Attributed Restriction, Disability, Social Handicap and Impaired Participation (HARDSHIP) questionnaire, a structured tool for epidemiological studies on headache disorders [10]. While the questionnaire was initially validated in adults, an adapted version for children and adolescents has been developed [11–14]. However, validation in the Italian pediatric population is still lacking.

This study aims to validate the Italian version of the HARDSHIP questionnaire as a first step to establish a comprehensive epidemiological registry, facilitating accurate monitoring of headache disorder prevalence, risk factors, and the development of targeted health policies.

Methods

Study design and population

REPICEF (Registro Epidemiologico delle Cefalee in età evolutiva - Epidemiologic Registry of Child and Adolescents Headaches) is a population-based registry investigating the prevalence and progression of headache disorders in school-aged children and adolescents in L’Aquila, central Italy. The study was approved by the Ethics Committee of the Abruzzo Region (protocol number 035018), and informed consent was obtained from parents or legal guardians of participants.

The study includes a cross-sectional phase, enrolling students aged 6–17 years from all schools in the city, which is testing the prevalence of headache disorders, and a prospective phase which will follow the trajectories of headache disorders in the same population. Data collection began in February 2024 and is ongoing, both to expand the baseline cohort and to enable longitudinal follow-up. Given that school attendance is mandatory in Italy up to the age of 16, the sample is considered representative of the general population in this age group. For this paper, headache diagnoses were attributed to all participants who reported at least one headache episode in the three months prior to questionnaire administration.

Inclusion and exclusion criteria

We included all children and adolescents aged 6–17 years attending selected schools in the city of L’Aquila. The selected schools - including “Istituto Comprensivo Gianni Rodari”, “Convitto Nazionale Domenico Cotugno”, “IIS Amedeo D’Aosta”, “IIS Ottavio Colecchi”, “Istituto Santa Maria degli Angeli”, and “Scuola Maestre Pie Filippini” - covered various educational levels, from elementary through high school, and comprised both public and private institutions. By involving schools from different educational levels and types, we achieved a sample that mirrors the demographic and socioeconomic distribution of school-aged children in the city of L’Aquila. We excluded children younger than 6 years or older than 17 years; additional exclusion criteria were lack of signed informed consent from participants’ parents or legal guardians and neurodevelopmental, psychiatric, or neurological disorders requiring educational support, including intellectual disabilities and reading or comprehension difficulties.

Questionnaire translation

The HARDSHIP was developed by Lifting The Burden to ensure a reliable assessment of headache disorders [10]. It includes 44 questions related to headache symptoms and their related burden. Questions about headache symptoms allow respondents’ classification into the following diagnoses of headache disorders: Migraine (Mig); Probable migraine (pMig); Tension-type headache (TTH); Probable tension-type headache (pTTH); Probable medication overuse headache (pMOH); Undifferentiated headache (UdH) for mild attacks lasting < 1 h, Headache on ≥ 15 days/month (H15+); and Undetermined headache for attacks and symptoms that do not meet any of the other diagnoses.

The translation of the HARDSHIP questionnaire followed the TRAPD method (Translation, Review, Adjudication, Pretest, and Documentation) [15]. This method consists of 5 different steps, namely translation, review, adjudication, pretest, and documentation [16]. All steps of the translation process were carefully documented.

Translation

Two translators (RO and AO) both proficient in English and native Italian speakers with experience in pediatric headache medicine independently translated the questionnaire from English to Italian (Fig. 1).

Fig. 1.

Translation of the HARDSHIP questionnaire

Review

The two preliminary translations were reviewed by the translators and an independent reviewer who was a native English and Italian speaker. For each question, the best wording was discussed to achieve a single translation.

Adjudication

The translation was compared and considered equal to the original (English) version by the independent reviewer of the previous phase.

Pretest

The translated questionnaire was pretested in the “pretest cohort.” During the pretest, 10 randomly chosen participants were asked to complete the questionnaire. Furthermore, participants were offered the opportunity to leave remarks about clarity and wording after each question. Once participants in the pretest cohort had completed the questionnaire, a phone interview was conducted with each of them.

Their interpretation of each of the questions and any perceived ambiguities were evaluated. Finally, participants were asked if they had any additional comments or remarks. All feedback that was collected during the pretest was reviewed by the two translators and the reviewer by repeating the first three steps of the TRAPD model until an agreement was reached on the revised final version of the translation. The final questionnaire is provided as supplementary material (Suppl.1).

Validation study

Four investigators (AO, CR, MDA, BG) administered the Italian version of the HARDSHIP questionnaire to every child and adolescent enrolled in the participating schools, irrespective of their history of headache.

For students in primary and lower secondary schools, the questionnaire was administered in small groups through guided completion, supported by a trained staff member. For high school students, the questionnaire was self-completed in the classroom, but always under the supervision of an investigator available to clarify any doubts. For this study, we only considered the HARDSHIP questions related to headache diagnostic criteria, while those related to headache impact were not considered.

Diagnoses based on the questionnaire were derived using the official HARDSHIP diagnostic algorithm, following its hierarchical sequence (H15+, pMOH, UdH, definite migraine, definite TTH, probable migraine, probable TTH), as developed by the Global Campaign against Headache. Then, a subset of participants was interviewed by telephone, together with their parents, by two expert clinicians (RO and SS), who were blinded to the HARDSHIP responses. Clinical diagnoses were made according to the ICHD-3 criteria [17]. To ensure diagnostic consistency, participants who did not meet the criteria for any ICHD-3-defined primary headache disorder were excluded from the comparison. Both children and parents contributed to the interview; however, clinicians prioritized the child’s self-report of symptoms, consistent with literature suggesting greater reliability in pediatric headache assessment. Agreement was considered for migraine, TTH, and pMOH, if the questionnaire diagnosis of pMOH had similar diagnostic criteria to the ICHD-3 diagnosis of medication overuse headache. HARDSHIP questionnaires were filled in on paper and then transcribed in the Research Electronic Data Capture (REDCap) framework hosted by the University of L’Aquila. Clinical diagnoses were directly inserted in the REDCap framework to minimize errors. Each participant was given a unique code that was used to match questionnaire-based with clinical diagnoses. Identifying information such as participants’ names and dates of birth were not inserted into the REDCap framework.

Statistical analysis

We reported descriptive data with numbers and proportions for discrete variables and with mean ± standard deviation (SD) for continuous variables. Sensitivity, specificity and positive (PPV) and negative (NPV) predictive values of the questionnaire for migraine, TTH, and pMOH, along with kappa values for agreement between the different diagnostic systems, were calculated. To explore whether age influenced diagnostic agreement, we repeated the analysis separately for children aged 6–11 years and adolescents aged 12–17 years. Definite and probable diagnoses of migraine and TTH as results of the HARDSHIP or of the clinical interview were considered together to evaluate agreement. In accordance with established recommendations for questionnaire validation, which suggest recruiting between 5 and 10 participants per item to achieve stable factor structures and reliable estimates, we determined a minimum target sample of 450 participants, based on the 44 items of the HARDSHIP questionnaire. Statistical analysis was performed using R, version 4.2.2 (The R Project for Statistical Computing, Vienna, Austria) and RStudio, version 2022.12.0 + 353.

Results

Study population

A total of 1023 children and adolescents were selected to take part in the project between February and December 2024. Twelve subjects were excluded from the study due to neurodevelopmental, psychiatric, or other neurological disorders, while 153 were excluded because their parents did not provide consent for participation. A total of 858 children and adolescents (453 males, 52.8%) aged 6–17 years (mean age 13.5 ± 3.0 years) were evaluated and included in the REPICEF cohort. Of these, the first 535 (62.4%) constituted the cohort for the validation of the HARDSHIP questionnaire (Fig. 2).

Fig. 2.

Selection process up to the validation of the HARDSHIP questionnaire

Questionnaire-based diagnoses

One hundred ninety-six participants (22.9%) had not experienced headaches in the past three months and thus were excluded. In subjects who experienced headache, based on questionnaire responses, pMIG was diagnosed in 175 (20.7%), pTTH in 144 (16.7%), UDH in 105 (12.2%), TTH in 86 (10.0%), migraine in 68 (7.9%), H15 + in 40 (4.6%), pMOH in 38 (4.4%), and undetermined headache in 6 (0.7%).

Clinical diagnoses and questionnaire validation

Clinical diagnoses included pMIG in 141 (26.4%), migraine in 107 participants (20.0%), TTH in 54 (10.1%), pTTH in 44 (8.2%), H15 + in 12 (2.2%), and pMOH in 3 (0.6%). One hundred seventy-four children and adolescents remained undiagnosed as they had not experienced headaches in the past three months. Table 1 shows the sensitivity, specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV), and kappa agreement for the comparison between questionnaire-based and clinical diagnoses. The sensitivity of the HARDSHIP questionnaire was low in the presence of high specificity for both migraine and TTH, while the sensitivity and specificity of the HARDSHIP questionnaire were both high for pMOH, Agreement between the HARDSHIP questionnaire and clinical diagnoses was moderate for migraine (κ = 0.432), fair for TTH (κ = 0.327), and low for pMOH (κ = 0.214). Table 2 reports the agreement between questionnaire and clinical diagnoses according to age groups. Concordance was higher for migraine among children (κ = 0.768) than among adolescents (κ = 0.361), while TTH showed greater agreement in adolescents (κ = 0.643) compared to children (κ = 0.359). Agreement for pMOH was low in both groups due to the very low prevalence of the condition.

Table 1.

Accuracy of HARDSHIP headache diagnosis compared to the gold standard clinical diagnosis: sensitivity, specificity, and agreement measures

Headache	Sensitivity	Specifity	PPV	NPV	Kappa (95% CI)
Migraine*	0.50	0.92	0.84	0.68	0.432 (0.354–0.510)
TTH*	0.60	0.79	0.39	0.90	0.327 (0.226–0.427)
pMOH	1.00	0.96	0.12	1.00	0.214 (−0.115–0.544)

Abbreviations: TTH Tension-type headache, pMOH Probable medication overuse headache, PPV Positive predictive value, NPV Negative predictive value

*Includes definite and probable diagnoses together

Table 2.

Diagnostic agreement between HARDSHIP questionnaire and clinical diagnoses stratified by age group

Diagnosis	Diagnostic Parameter	Children (6–11 y)	Adolescents (12–17 y)
Migraine	Sensitivity	0.86	0.74
	Specificity	0.96	0.84
	PPV	0.75	0.31
	NPV	0.98	0.97
	Kappa (95% CI)	0.768 (0.517-1.000)	0.361 (0.256–0.467)
TTH	Sensitivity	0.38	0.72
	Specificity	0.93	0.96
	PPV	0.62	0.63
	NPV	0.83	0.97
	Kappa (95% CI)	0.359 (0.063–0.654)	0.643 (0.520–0.765)
pMOH	Sensitivity	0.00	0.13
	Specificity	1.00	1.00
	PPV	0.00	1.00
	NPV	0.98	0.96
	Kappa (95% CI)	0.000 (0.000–0.000)*	0.222 (0.000-0.432)

Abbreviations: TTH Tension-type headache, pMOH probable medication overuse headache, PPV Positive predictive value, NPV Negative predictive value

*No cases according to clinical diagnoses

Discussion

Our results highlight the challenges in accurately classifying headache disorders in the pediatric population. With relatively low sensitivity but high specificity compared to clinical diagnoses, the Italian version of the HARDSHIP questionnaire may serve as a useful screening tool for primary headache disorders in general settings, although confirmation by expert clinicians remains necessary. The limited diagnostic agreement observed may reflect two key factors: first, differences in symptom reporting between questionnaire-based assessments and clinical interviews; second, structural differences between the HARDSHIP algorithm and the ICHD-3 diagnostic framework, particularly the inclusion in HARDSHIP of categories such as undifferentiated headache, which are not captured by ICHD-3 criteria. It might be useful to detect the prevalence of primary headache disorders in the pediatric population and maintain surveillance over large populations in the long term, thereby serving epidemiological purposes more effectively than the case series of specialized headache centers, which only take care of the most severe cases of headache disorders. The agreement between questionnaire-based and clinical diagnoses in our validation study was moderate for migraine and fair for TTH, while it was low for pMOH, despite the high sensitivity and specificity of the questionnaire for this diagnosis, given the low prevalence of pMOH among children and adolescence. The agreement between diagnoses established by the questionnaire and those made by physicians aligns with validation results from other languages. While in adult Indian population [16] the kappa values were 0.43 for both migraine and TTH, in another Indian study [17] they were 0.46 for migraine and 0.39 for TTH. The observed variations in kappa values across different studies may be partly attributed to methodological differences, sample characteristics, and healthcare system variability. However, deeper factors should also be considered, including cultural differences in symptom expression, linguistic challenges in translating key terms, particularly those describing subjective experiences such as photophobia or phonophobia, and variability in the application and interpretation of the diagnostic gold standard across settings and countries. In the pediatric population, a Lithuanian [9] and a Danish [18] study validated the HARDSHIP questionnaire via a test-retest method instead of comparing questionnaire-based with clinical diagnoses and therefore they did not calculate sensitivity, specificity, or predictive values with respect to a gold standard.

The main strength of the present study is the validation of the HARDSHIP questionnaire against the diagnoses provided by clinical experts, which are currently the most accepted reference standard in pediatric headache diagnosis. Besides, we performed the validation in a well-defined cohort of children and adolescents from several schools in a city, representing the general population of 6–17 years of age in Western countries. However, this study also has some limitations. One major limitation concerns the generalizability of the findings, as the study population consists primarily of children and adolescents from a single Italian region, which may not fully represent broader, more diverse populations. Additionally, the inherent limitations of the questionnaire should be considered, as it resulted in a high number of probable diagnoses and relatively few definitive classifications, which may impact its diagnostic precision. Another limitation relates to the construct validity of the questionnaire. Since this study did not assess dimensions beyond headache symptoms, such as comorbidities and quality of life, our findings do not provide insights into the broader impact of headache disorders, a limitation also noted in similar studies, including the Lithuanian validation of the HARDSHIP questionnaire [11]. Although the present study focused exclusively on the diagnostic items of the HARDSHIP questionnaire, it is worth noting that the instrument also includes a section addressing the impact of headache on daily functioning. This component provides a multidimensional assessment, though less commonly used in clinical practice compared to validated disability scales such as PedMIDAS and HIT-6. These tools offer more detailed and standardized measures of headache-related disability and are frequently employed to monitor longitudinal outcomes or assess treatment efficacy. Future research may consider comparing the impact-related items of HARDSHIP with these established tools in pediatric populations. Furthermore, the study may be affected by selection bias, as children and adolescents with significant cognitive or neurological impairments, who required special education support, were excluded.

Conclusions

The Italian validation of the HARDSHIP questionnaire provides a tool for epidemiological surveillance of headache disorders in the pediatric population. By establishing a validated approach for headache assessment, this study lays the groundwork for a longitudinal assessment of the incidence and prevalence of headache disorders among children and adolescents, with potential improvements in early diagnosis and in treatment strategies.

Authors’ contributions

AO, RO, and CR designed the study, performed statistical analysis, and revised the article for intellectual content; AO and RO translated the questionnaire. AO, RO and CR coordinated data collection and wrote the first draft of the manuscript; SS revised the manuscript; all authors participated in data collection and approved the final version of the manuscript.

Funding

Open access fee for this paper was provided by a grant from the Italian Ministry of Health - PRIN 2022 (project code 20228Y8S5F).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

AO is member of the Junior Editorial Board of TJHP; RO reports personal fees from AbbVie, Bayer, Eli Lilly, Lundbeck, Novartis, Organon, Pfizer, and Teva and grants from the Italian Ministry of Health; he is member of the Editorial Board of TJHP, Frontiers in Neurology, Arquivos de Neuropsiquiatria, and Confinia Cephalalgica; SS reports personal fees from Novartis, Uriach, Royalties, Abbott, Allergan-AbbVie, AstraZeneca, Boheringer, Eli Lilly, Lundbeck, Novartis, NovoNordisk, Pfizer, Teva. Consulting fees, Abbott, Allergan-AbbVie, AstraZeneca, Boheringer, Eli Lilly, Lundbeck, MedScape, Novartis, NovoNordisk, Pfizer, Teva. Expert testimony, Abbott, Allergan-AbbVie, AstraZeneca, Boheringer, Eli Lilly, Lundbeck, Novartis, NovoNordisk, Pfizer, Teva, Bayer, Medtronic, Starmed, Bristol-Myers-Squibb, Daiichi-Sankyo. Patents, Allergan-AbbVie, AstraZeneca, Eli Lilly, Lundbeck, Novartis, Pfizer, Teva. Leadership, President-elect European Stroke Organization, Editor-in-Chief Cephalalgia and Cephalalgia Reports, assistant editor for Stroke; CR, VT, MTA, BG, MDA, CC, MER, SF, RC declares no conflicts of interest.