PedMIDAS-Based Scoring Underestimates Migraine Disability on Non-School Days


  • Geoffrey L. Heyer MD,

    Corresponding author
    1. Departments of Pediatrics and Neurology, Nationwide Children's Hospital, The Ohio State University, Columbus, OH, USA
    • Address all correspondence to G.L. Heyer, Departments of Pediatrics and Neurology, Nationwide Children's Hospital and The Ohio State University, 700 Children's Drive, ED-5, Columbus, OH 43205, USA.

    Search for more papers by this author
  • Kelsey Merison MD,

    1. Departments of Pediatrics, Nationwide Children's Hospital, Columbus, OH, USA
    Search for more papers by this author
  • Sean C. Rose MD,

    1. Departments of Pediatrics and Neurology, Nationwide Children's Hospital, The Ohio State University, Columbus, OH, USA
    Search for more papers by this author
  • Sara Q. Perkins,

    1. Center for Human Psychophysiology, Miami University, Oxford, OH, USA
    Search for more papers by this author
  • JoEllen M. Lee CNP,

    1. Departments of Pediatrics and Neurology, Nationwide Children's Hospital, The Ohio State University, Columbus, OH, USA
    Search for more papers by this author
  • William C.L. Stewart PhD

    1. Departments of Pediatrics, Nationwide Children's Hospital, Columbus, OH, USA
    2. Statistics Department, The Ohio State University, Columbus, OH, USA
    Search for more papers by this author

  • Funding: This work was supported by a Nationwide Children's Hospital Intramural Grant [Grant #278311; GL Heyer].
  • Conflicts of Interest: None.



The aim of this study is to compare daily Pediatric Migraine Disability Assessment (PedMIDAS)-based scores for headaches occurring on school days vs non-school days and during the school year vs the summer holiday.


The PedMIDAS is the only instrument validated to assess migraine disability among school-aged children. However, the PedMIDAS may underestimate disability during prolonged school holidays.


In a prospective cohort study, migraine patients aged 10–18 years completed a 90-day Internet-based headache diary. For each headache day, they answered PedMIDAS-based questions and rated their headache intensity (scale 1–10). PedMIDAS-based scores, headache intensity ratings, and relative headache frequencies were compared for school days vs non-school days and for the school year vs the summer holiday.


Fifty-two patients completed 4680 diary entries comprising 984 headache days. The headache frequencies and intensity ratings did not differ between time periods. However, the mean headache disability scores (as measured from PedMIDAS-based questions) were significantly different for school days (0.85) compared to non-school days (0.45), P < .001, and for the school year (0.73) compared to the summer holiday (0.46), P < .016.


Given similar headache intensities and frequencies, daily PedMIDAS-based scores significantly underestimate headache disability on non-school days. Accordingly, PedMIDAS scoring during the school year may not be comparable to assessments done during the summer holiday. These potential differences must be considered when using the instrument as an outcome measure for clinical trials.


Migraine Disability Assessment


Pediatric Migraine Disability Assessment

Migraine is a common form of primary headache that often begins during the early school-age years.[1, 2] The disability caused by migraines can lead to impairments in a child's daily activities and school performance and can adversely impact quality of life.3-5

The Pediatric Migraine Disability Assessment (PedMIDAS) is the only validated clinical tool available to estimate migraine disability in the school-aged child.[5] The instrument consists of 6 questions addressing school absence, poor functioning during school, and disruption of home and social/recreational activities over a 3-month recall period. In the clinical setting, the PedMIDAS can be helpful in assessing a patient's migraine burden and response to therapy. Some researchers advocate its use as an outcome measure in clinical trials.[5]

While the PedMIDAS is comparable to the adult Migraine Disability Assessment (MIDAS),[6] differences exist between school-aged children and adults. While most adults have year-round work and/or household duties, most school-aged children have a prolonged interruption in school attendance each summer. Fifty percent of the PedMIDAS questions address school-related disability, so disability scoring should be systematically higher on school days than on non-school days. Accordingly, scores from the 3-month PedMIDAS could vary as a function of the date(s) of administration relative to the school year. These potential scoring inconsistencies must be identified and addressed if the instrument is to be used satisfactorily as an outcome measure in clinical research.

The aim of the current study was to compare headache frequency, PedMIDAS-based headache disability, and headache intensity for school days vs non-school days and, separately, for the school year vs the summer holiday. An Internet-based headache diary was used to track all study variables.



We conducted a prospective study of 52 patients with migraine or probable migraine over an 11-month period of time from December 2011 to October 2012. Each patient completed a 90-day Internet headache diary which incorporated PedMIDAS-based questions, revised to address headache disability for each headache day. Headache frequencies, disability scores, and intensity ratings were compared for school vs non-school days and for the school year vs the summer holiday.


This study was approved by the Institutional Review Board at Nationwide Children's Hospital. Written informed consent (parents and subjects 18 years of age) and assent (subjects <18 years) were obtained in all cases.

Patient Population

Patients ranged in age from 10–18 years, and all had clinical histories consistent with episodic migraine (with or without aura) or probable migraine based on International Headache Society criteria.[7] We included patients with 1–15 headaches monthly. Migraine patients who also had episodic tension-type headaches were not excluded provided that their migraines occurred ≥1 day per month, and the frequency of all combined headaches remained ≤15 days per month. The sample size was chosen empirically (goal of n = 50) to include adequate variations in patient age, headache frequency, and migraine disability.

Internet Headache Diary

The Internet diary has been previously described.[8] The original intent of the diary design was to evaluate headache recall among youth and their parents. Briefly, daily diary entries included 11 questions and an optional comment section. The first question asked “Did you have a headache today?” Questions 2 to 7 mirror PedMIDAS questions[5] but were modified to address disability for each daily diary entry. Questions 2 to 4 addressed missing school (Q2), missing partial school days due to leaving early or arriving late (Q3), and functioning at less than half ability in school (Q4) because of a headache. Question 5 asked if activities at home such as homework or chores were affected by headache. Questions 6 and 7 addressed missed participation in social or recreational activities (Q6) and functioning at less than half ability during activities because of headache (Q7). In keeping with the PedMIDAS structure, patients could not choose more than one form of disability for school or for social activities for a given headache day. For example, if Q2 (“missed school”) was selected, then Q3 and Q4 were automatically blocked. Question 8 provided a headache intensity rating scale that ranged from 1 to 10. Questions 9 to 11 addressed medicine compliance.

Study Protocol

Patients were asked to complete a diary entry each day. Study investigators had an administrative login feature that allowed review of all daily diary entries upon submission and monitoring of daily compliance. Daily e-mail reminders were sent to parents and patients when entries were missed. Families were contacted by telephone after 5 consecutive missed days. Patients were asked to complete all missed entries by describing headache disability and intensity in the comment section of the subsequent entry or by relaying information to the study coordinator by e-mail or telephone.

A disability score was calculated for each headache day. The score ranged from 0 to 3 based on the sum of affirmative responses to three PedMIDAS disability categories: school (Q2-Q4); home activities (Q5); and leisure/recreational activities (Q6-Q7). Patients distinguished school days from weekends and holidays when answering school-related questions (ie, Did you miss school today because of a headache?) as “yes,” “no,” or “weekend or school holiday.” Weekend and holiday designations were confirmed by comparing the date-stamped diary entry to the school-district calendar. The school year was defined as all school days (including weekends and school holidays) beginning from the first school day through the last school day of the calendar year. The summer holiday comprised all calendar days not included in the school year.


To assess the evidence for systematic differences in headache disability, intensity, and frequency, we tested the null hypothesis of no difference between means for school days vs non-school days and for the school year vs the summer holiday. The 90-day observation period contained weekdays during the school year, weekends during the school year, and (for n = 32 patients) days during the summer holiday. To account for inter-group correlation (ie, some subjects were observed across multiple time periods), we used the nonparametric bootstrap procedure.[9] The observed data were resampled 10,000 times with replacement. For each resample (aka bootstrap replicate), the correlation was preserved, and the mean difference for disability, intensity, and the presence/absence of headaches was computed. The standard bootstrap confidence intervals were constructed, and then inverted, to obtain two-sided P-values. For each of these tests, significance was assessed at the α = 0.025 level in accordance with the standard Bonferroni correction to ensure a family-wise error rate that is less than 5%. All statistical analyses were performed using R (Version 3.0.1, R Foundation for Statistical Computing, Vienna, Austria).


Fifty-two subjects completed 90 consecutive diary entries for a total of 4680 diary days. Seventy-eight patients were initially recruited: 12 never began and 14 were removed during the study period (10 for noncompliance, 3 for daily headaches, and 1 for loss of home Internet access). There were no statistical differences between the 52 study completers and the 26 noncompleters in terms of age, sex, migraine vs probable migraine, or initial patient-reported headache frequencies (data not shown). Of the 4680 diary days, 3215 (68.7%) were completed via same-day diary entries. The remaining 1465 (31.3%) missed days were completed through the diary comment section of subsequent entries and through responses to e-mail or telephone reminders. All missed diary entries were completed within 8 days of the expected entry.[8]

Patients reported headaches on 984 (21%) of 4680 diary entries. For all 984 headache days, the mean headache intensity rating was 5, and the mean disability score was 0.62. Table  summarizes the relative headache frequencies, mean headache intensity ratings, and mean disability scores for each studied time period. Headache frequency and intensity ratings did not differ between school days and non-school days or between the school year and the summer holiday (Table ). However, using daily PedMIDAS-based questions, headache disability scores were significantly lower for non-school days. School-day disability (mean 0.85) differed from weekends and holidays (mean 0.45), P value < .001; and school-year disability scores (mean 0.73) differed from summer-holiday scores (mean 0.46), P value <.016. Figure a,b depicts mean disability score as a function of headache intensity rating for each time period of interest.

Figure –.

Irrespective of headache intensity, the mean disability scores on school days exceeded the mean disability scores on non-school days (a). Similarly, the mean disability scores during the school year exceeded the mean disability scores during the summer holiday, with the single exception of headaches with 9/10 intensity ratings (b).

Table –. Headache Frequencies, Intensity Ratings, and PedMIDAS-Based Disability Scores
 Total Diary DaysTotal HA DaysPercent HA Days [CI]Mean HA Intensity Ratings [CI]Mean HA Disability Scores [CI]
  1. aSchool days vs non-school days: headache frequency, P = .91; headache intensity ratings, P = .23; headache disability scores, P < .001.
  2. bSchool year vs summer holiday: headache frequency, P = .93; headache intensity ratings, P = .48; headache disability scores, P < .016.
  3. Statistically significant differences are emboldened.
  4. HA = headache; CI = 95% confidence intervals.
School daysa183341022.4 [0, 54.2]5.2 [4.6, 5.8]0.85 [0.81, 0.89]
Non-school days284757420.2 [4.5, 35.9]4.9 [4.4, 5.4]0.45 [0.41, 0.49]
School-year daysb277158121 [0, 47.9]5.1 [4.5, 5.7]0.73 [0.69, 0.77]
Summer-holiday days190940321.1 [3.1, 39.1]5.0 [4.3, 5.7]0.46 [0.42, 0.50]

Given that the maximum daily disability scores differ between school days (maximum of 3) and non-school days (maximum of 2), we conducted a post-hoc analysis of disability scores scaled by their respective maximum values. The mean disability proportion for school days (33.6%) was not statistically different from the mean disability proportion for non-school days (24.1%), P value .104. The mean difference in disability proportions was 9.5% (95% confidence interval 0, 0.2275).


The results of our study indicate that daily PedMIDAS-based disability scoring differed based on the presence or absence of school that day. As the headaches themselves did not clearly differ, headache frequency and intensity remained similar across the studied time periods, it is the tool used (ie, the PedMIDAS-based questions) that determined the change in disability scores.

Our method of daily headache disability scoring parallels the PedMIDAS which sums daily disability over a 3-month recall period. Accordingly, as daily disability scores vary in relation to the presence or absence of school on a given day, PedMIDAS scores also may vary relative to the ratio of school days vs non-school days during the 3-month recall period. Using elements of the PedMIDAS modified for daily assessment, a school day can be scored as high as 3; while a non-school day can be scored no higher than 2. In a post-hoc analysis, we scaled the daily disability scores by their maximum possible values, and the comparison of proportions was no longer statistically significant. The lack of a significant difference between the scaled disability proportions and the similar headache intensity ratings between school days and non-school days each suggest that the perceived headache disability did not differ between time periods. Rather, the difference in disability scores represents a ceiling effect related to the PedMIDAS instrument.

While the PedMIDAS and MIDAS are similar in terms of the types of disability measured, adults are less likely to have daily work or household work interrupted by a prolonged summer holiday, so the effects of holidays on MIDAS scores may be less important. Weekends and brief holidays during the adult work year should average out over a given 3-month recall period. Consequently, assuming that headache burden does not change, MIDAS results would remain relatively consistent regardless of the time of year. In contrast, PedMIDAS scores obtained midway through the school year may be significantly higher than scores obtained near the end of summer, even if headache frequencies and intensities remained similar between assessments. Unfortunately, in a clinical trial using the PedMIDAS as an outcome measure, seasonal inconsistencies in scoring could produce false-positive or false-negative treatment effects.

There are several potential study limitations. We used a daily measure of headache disability and PedMIDAS questions modified for daily use rather than the PedMIDAS itself which can affect the validity of the resulting scores. Inaccuracy in 3-month recall limits the direct comparison of seasonal PedMIDAS scoring. Also, we did not differentiate between migraine disability and disability due to tension-type headaches. Patients with fewer migraine headaches compared to tension-type headaches could have smaller differences between school and non-school disability. Lastly, our patient population was highly selected. Selection factors include referral to a headache clinic, home Internet access, age, and diary compliance, all of which minimize our ability to generalize results.


Daily PedMIDAS-based disability scores were significantly lower for non-school days vs school days and for the summer holiday vs the school year, while other headache factors did not differ. The number of non-school days during the 3-month PedMIDAS recall period could lead to scoring inconsistencies, particularly when comparing headache disability during summer months and school months. These potential scoring inconsistencies must be considered when using the instrument as an outcome measure in year-round clinical trials.


We would like to thank Aggie LeGros, RN (Nationwide Children's Hospital) for her daily coordination of this study. We would also like to acknowledge Drs. E. Steve Roach and Kathi Kemper (Nationwide Children's Hospital and The Ohio State University) for their helpful recommendations regarding manuscript preparation.

Statement of Authorship

Category 1

  • (a)Conception and DesignGeoffrey L. Heyer
  • (b)Acquisition of DataGeoffrey L. Heyer; JoEllen M. Lee
  • (c)Analysis and Interpretation of DataGeoffrey L. Heyer; William C. L. Stewart; Kelsey Merison; Sean C. Rose; Sara Q. Perkins

Category 2

  • (a)Drafting the ManuscriptGeoffrey L. Heyer
  • (b)Revising It for Intellectual ContentGeoffrey L. Heyer; Kelsey Merison; Sean C. Rose; Sara Q. Perkins; JoEllen M. Lee; William C. L. Stewart

Category 3

  • (a)Final Approval of the Completed ManuscriptGeoffrey L. Heyer; Kelsey Merison; Sean C. Rose; Sara Q. Perkins; JoEllen M. Lee; William C. L. Stewart