Original Article
Four methods of estimating the minimal important difference score were compared to establish a clinically significant change in Headache Impact Test

https://doi.org/10.1016/j.jclinepi.2005.05.010Get rights and content

Abstract

Objective

To estimate the smallest decrease in Headache Impact Test (HIT) scores that reflects meaningful clinical change among patients with chronic daily headache (CDH).

Study Design and Setting

We applied four methods of estimating the minimum important difference (MID) to data from 71 patients with CDH who participated in a clinical trial. The HIT was administered at baseline and at the 6-week follow-up assessment. Patients were considered to have experienced meaningful improvement if they reported that their headache condition was “somewhat better” or “much better” at the 6-week follow-up.

Results

Mean HIT scores at baseline and 6 weeks for all patients were 64.5 (standard deviation SD = 6.0) and 62.6 (SD = 5.7), respectively. HIT scores decreased 3.7 (SD = 4.4) and 1.4 (SD = 3.6) units, respectively, among patients who reported “somewhat better” change and those who reported no change at 6 weeks. Estimates of the MID of the HIT ranged from −2.7 to −2.3.

Conclusions

The method that we judge to be most valid estimated the MID of the HIT at −2.3 units (95% confidence interval = −4.3, −0.3). This suggests that a between-group difference in HIT change scores of 2.3 units over time among patients with CDH reflects improvement in patients' headache condition that may be considered clinically significant.

Introduction

Chronic daily headache is a clinical syndrome characterized by the presence of headache ≥15 days per month [1]. Surveys of the general adult population suggest that the prevalence of chronic daily headache is ∼3–5% [2], [3]. Few standardized instruments to assess meaningful clinical change among patients with chronic daily headache are available, and no single instrument is widely accepted for this purpose. The paucity of standardized outcome measures for chronic daily headache makes it difficult to assess clinical change among individual patients over time, or to compare the efficacy of treatments for chronic daily headache.

The Headache Impact Test (HIT; also HIT-6) is a standardized and validated questionnaire that may be useful for assessing clinical change over time among patients who experience frequent headaches. The questionnaire comprises six items pertinent to persons who experience frequent headaches. This instrument was found to be reliable and valid among a heterogeneous population of headache sufferers who completed a survey administered over the Internet [4], [5]. The HIT also appears to be responsive to change in clinical status over time; headache sufferers who reported improvement in the degree to which headaches interfered with daily activities demonstrated a concomitant decrease in mean HIT scores of about three units [4]. However, the performance of the HIT among patients with chronic daily headache is unknown.

Our primary objective was to estimate the minimal important difference (MID) of the HIT. The aim is to help clinicians interpret changes in HIT scores among individual patients and to help investigators interpret differences in changes in HIT scores between two or more groups in the context of clinical trials designed to evaluate the efficacy of treatments for chronic daily headache. Our secondary objective was to compare various methods of estimating the MID of standardized questionnaires.

Section snippets

Methods

We obtained data from headache patients who participated in a randomized clinical trial comparing usual medical care to usual medical care plus a series of 10 acupuncture treatments over a period of 6 weeks. Patients were recruited from a headache-specialty clinic located within an academic medical center. Every patient had chronic daily headache, defined as the presence of headache on ≥15 days in the month prior to enrolling in the clinical trial. We obtained complete follow-up data from 71 of

Results

Patients' ages ranged from 19 to 83 years, with a mean of 46. Fifty-seven (80%) were female, 66 (93%) were white, 43 (61%) were married, 38 (54%) had obtained a college degree, and 40 (56%) were employed. Patients reported a mean of 24.2 days (standard deviation SD = 5.8) with headache in the month prior to enrolling in the study, with an average pain severity of 6.4 (SD = 2.0) on an 11-point scale. There were no significant differences in baseline measures between the two treatment arms. Mean

Discussion

Our findings suggest that the value of HIT change scores that reflects meaningful clinical change differs depending on whether a single individual is being followed over time (i.e., within-person change) or whether two groups of patients are being compared, such as in the setting of a clinical trial (i.e., between-group difference). Mean HIT scores decreased by 1.4 units (P < .05) among patients who reported no clinical change during the 6-week study period. Statistical significance of this

Summary

We applied all of the methods of estimating the MID of a health-related quality of life instrument identified in the published literature for which we had sufficient and appropriate data to analyze. The methods described and the data used in this investigation, although imperfect, reflect the current state of the science of MID research. Our findings suggest that the within-person MID of the HIT for clinical improvement is ∼3.7 units and that the between-groups MID is ∼2.3 units. The

Acknowledgments

R.R.C. is supported by a Mentored Research Career Development Award, no. K23-AT-001194 from the National Center for Complementary and Alternative Medicine. Funding for this investigation was also provided by NIH grant no. RR00046 and an investigator-initiated grant from GlaxoSmithKline.

References (20)

There are more references available in the full text version of this article.

Cited by (120)

View all citing articles on Scopus
View full text