What we found
A published trial of 2,702 women. The scoring was blind to the result.
Letter A Got more back
Surveys sent back26%
The letterRewritten
Aperture Score ranked itHigher
Letter B
Surveys sent back21%
The letterStandard
Aperture Score ranked itLower
What the Aperture Score caught: Letter A said what to do, and by when.
The rewritten letter changed about a dozen things at once, a three-week deadline among them, and the trial could not say which one did the work. One trial is one trial. When we ran the same blind test on 18 more published trials, the ranking did no better than chance at picking the version that got more people to act. The Aperture Score measures how a piece is built for a frightened reader. It does not predict how many people will reply.
McBride E, Mase H, Kerrison RS, Marlow LAV, Waller J. BMC Medical Research Methodology, 2021.
Not reading level
Thirty-seven cancer-center documents were given an Aperture Score and, separately, scored with two readability formulas.
9%
of the Emotional Safety score is explained by reading grade.
5%
of the Action Clarity score is explained by reading grade.
37of 37
documents failed the same check: breaking a complicated procedure into numbered steps.
A trial that points the same way
In a 2003 randomized trial across 44 institutions, an easy-to-read consent form lowered patients' anxiety about consenting and raised their satisfaction. Their understanding didn't change.
Coyne CA and colleagues, Journal of Clinical Oncology, 2003. A consent form for a clinical trial, and someone else's research, not ours.
AHRQ, TalkingQuality: readability formulas ignore most of what drives comprehension.
The Aperture Score reads the material, not the patient. No study yet ties a score to a medical outcome.
Where the science comes from
We have read and catalogued more than 380 published studies. These 14 bear most directly on the five dimensions.
Earl and Albarracín Health Psychology, 2007
Meta-analysis, 150 treatment groups, HIV prevention
Fear-inducing arguments raised people’s sense of risk but lowered what they knew. Resolving the fear through counseling and testing raised knowledge.
Büchter and colleagues BMC Medical Informatics and Decision Making, 2014
Systematic review and meta-analysis, 10 randomized trials
Words like “common” and “rare” led people to overestimate the risk of side effects more than numbers did.
Petrie and colleagues BMJ, 2007
Randomized trial, 92 cardiology patients
Explaining beforehand what a normal result would mean raised reassurance, and fewer of those patients still reported chest pain a month later.
Sheridan and colleagues Journal of Health Communication, 2011
Systematic review, 38 studies, low health literacy
Putting the essential information first, or by itself, improved comprehension in one or a few studies.
Jansen and colleagues Journal of Clinical Oncology, 2008
260 newly referred cancer patients, recorded consultations
The more information patients were given about prognosis, the less information they recalled.
Laws and colleagues PLoS One, 2018
189 specialty-care visits
A week later, patients accurately recalled 49 percent of what was decided, unprompted. More items in a visit meant worse recall for patients with less schooling.
Rey and colleagues Educational Psychology Review, 2019
Meta-analysis, 56 investigations, in learners
Breaking material into meaningful segments improved learning.
Noetel and colleagues Review of Educational Research, 2022
Overview of 29 reviews, in learners
Good design mattered more when the material was complex, and when the pace was set for the learner, as in a lecture.
Gibson and Powell Thorax, 2004
Review of 26 randomized trials, asthma
Individual written plans that said when and how to step up treatment, in two to four action points, consistently improved outcomes.
Wolf and colleagues Medical Care, 2011
Controlled trial, not randomized, 500 patients, medicine labels
Explicit, plain wording was read correctly more often. Adding a picture made it slightly worse.
Sheeran and Orbell Health Psychology, 2000
Experiment, 114 women
Women who wrote down when, where and how they would book went to cervical screening more often: 92 percent against 69.
Traeger and colleagues JAMA Internal Medicine, 2015
Meta-analysis, 14 trials, 4,872 patients with acute low back pain
Patient education in primary care modestly increased reassurance, in the short term and the long term.
Pincus and colleagues Pain, 2013
Systematic review, 16 cohort studies
Explanation and education went with better outcomes. Empathy and rapport gave mixed results, and in three high-quality studies went with worse symptoms later.
Sep and colleagues Patient Education and Counseling, 2014
Randomized experiment, 50 healthy women watching a filmed bad-news consultation
A promise not to abandon the patient lowered physical arousal and raised recall.
Eysenck, Derakshan, Santos and Calvo Emotion, 2007
Theory paper
Attentional control theory: anxiety weakens goal-directed attention and pulls it toward threat.
Moran Psychological Bulletin, 2016
Meta-analysis, 177 samples, 22,061 people
Anxiety went with lower working memory.
Healthy People 2030 U.S. Department of Health and Human Services
National health objectives
Organizations share the responsibility for making health information usable.
None of these studies tested the Aperture Score. They are the published findings each dimension is built on. Three are thinner than the others: no trial has yet changed only the order of a patient document, the segmenting trials are in learners, not patients, and the reassurance trials are spoken or filmed, not written.