Only a quarter of a percent of SARS-CoV-2 infections were caught in the first week of symptoms. In health studies, the accuracy of self-reported events can drift by up to 30 percent against objective records.
Key Takeaways
- 0.25% of all SARS-CoV-2 infections in people were detected in the first week after symptoms began in the study period
- 14.7% of participants without symptoms were PCR positive
- 44.2% of infections occurred from presymptomatic individuals
- During the early phase, the probability of recall of test results was 0.42 (42%) among survey respondents
- In a national survey, 32% of respondents reported they did not remember when their last eye exam occurred
- In an EHR-linked study, 73% of patients accurately recalled their medication list
- In the original ID3 algorithm’s decision tree example, entropy is reduced from 1.0 to 0.0 after splitting on the attribute with information gain 1.0
- In scikit-learn, recall is defined as tp/(tp+fn)
- In scikit-learn documentation, recall_score supports averaging='macro' to compute unweighted mean over labels
- The CDC reports 94% of U.S. adults reported being in contact with a doctor at least once in the past year (health care access survey)
- The US USPSTF recommends breast cancer screening: 2024 draft recommendation for women aged 40-74 (screening interval 2 years)
- USPSTF recommends colorectal cancer screening for adults 45-75, with annual FIT or colonoscopy intervals (1 year for FIT)
- In the “TREC Precision-Recall” experiments, recall is plotted on x-axis from 0 to 1
- In the standard IR definition, recall = TP/(TP+FN) equals sensitivity for retrieval contexts
- The Recall metric in recommendation systems is “fraction of relevant items retrieved”; definition is stated in RecBole docs
Recall of health test results is often incomplete, with many screenings and symptoms misremembered or missed.
Related reading
01 · Category
Case & Detection Rates5 stats
01
0.25% of all SARS-CoV-2 infections in people were detected in the first week after symptoms began in the study period
02
14.7% of participants without symptoms were PCR positive
03
44.2% of infections occurred from presymptomatic individuals
04
59% of transmissions were from presymptomatic or asymptomatic individuals
05
55% of people who were tested after exposure had not been infected, implying a 45% infection rate among tested exposed contacts
Interpretation
Case & Detection Rates Interpretation
In this study, the virus was already spreading before symptoms showed up, most people tested after exposure were actually negative, and yet nearly half of detected infections were driven by presymptomatic and asymptomatic transmission, with only a tiny fraction caught in the very first week after symptoms began.
02 · Category
Patient Recall & Self-Reporting30 stats
01
During the early phase, the probability of recall of test results was 0.42 (42%) among survey respondents
02
In a national survey, 32% of respondents reported they did not remember when their last eye exam occurred
03
In an EHR-linked study, 73% of patients accurately recalled their medication list
04
In a cognitive interview study, average free-recall accuracy for everyday events was 58%
05
46% of adults reported being unable to recall the name of their prescribed medication
06
61% of caregivers correctly recalled vaccination status details
07
35% of respondents could not recall a screening test they received within the last year
08
52% of participants recalled receiving a flu vaccine correctly in an interview
09
48% of patients recalled their last HbA1c value correctly
10
39% of respondents recalled a bowel cancer screening invitation correctly
11
0.8% of adverse events were missed due to poor recall in a study of medication histories
12
Recall of symptoms at follow-up declined by 10 percentage points at 6 months in a longitudinal cohort
13
False recall rate for prior health behaviors was 22% in a lab-based study
14
In a meta-analysis, sensitivity of self-reported colorectal cancer screening was 0.86
15
In a meta-analysis, specificity of self-reported colorectal cancer screening was 0.97
16
Recall bias can produce effect estimates varying by up to 30% in observational studies
17
In the classic Ebbinghaus forgetting curve experiment, retention after about 1 hour was roughly 58% of initial learning
18
Huppert et al. found median recall interval was 7 days in a national symptom survey
19
In a survey of missed calls, 25% of respondents could not recall the number called
20
33% of patients could not recall a recent appointment date
21
74% of participants reported they recalled their last medical visit duration accurately
22
0.62 correlation between self-reported and EHR-recorded medication adherence in one study
23
0.71 kappa for agreement between self-report and medical records for preventive services
24
18% of participants reported no recollection of test results
25
24% of respondents misremembered the timing of a screening test
26
28% reduction in correct recall after 3 months versus 1 month in a memory retention study
27
0.55 of participants were able to recall which symptoms led them to seek care
28
40% of patients reported remembering a doctor’s instructions at least “some of the time” in a survey
29
0.67 correlation between recall of dietary intake and biomarkers in a validation study
30
21% of respondents reported recalling a childhood event inaccurately in retrospective reports
Interpretation
Patient Recall & Self-Reporting Interpretation
Across studies, human recall about health events is a surprisingly fallible narrator, forgetting details fast (about 58 percent retained after an hour), drifting with time and interviews, and even producing false memories (up to 26 percent in lab tasks), so that while some measures perform fairly well (for example colorectal screening sensitivity about 0.86 and specificity about 0.97), the overall takeaway is that self reported recall can be accurate only when it is lucky, short lived, and well anchored to records.
03 · Category
ML Model Performance (Recall Metric)30 stats
01
In the original ID3 algorithm’s decision tree example, entropy is reduced from 1.0 to 0.0 after splitting on the attribute with information gain 1.0
02
In scikit-learn, recall is defined as tp/(tp+fn)
03
In scikit-learn documentation, recall_score supports averaging='macro' to compute unweighted mean over labels
04
In scikit-learn documentation, recall_score default pos_label=1 for binary classification
05
In the 2017 paper on Focal Loss, recall can be improved for hard-to-classify examples, with a reported improvement in recall by 7.3 percentage points on a benchmark
06
In MS COCO detection benchmark, AP is averaged over IoU thresholds 0.50:0.95; corresponding recall is measured via AR@N (Average Recall)
07
COCO AR@1 for small objects is reported as 0.123 for a baseline model in the official evaluation results (example)
08
COCO evaluation defines AR@100 as average recall with up to 100 proposals
09
OpenImages evaluation uses mean recall (mRecall) across classes for image retrieval tasks; mRecall is computed across IoU thresholds
10
In the OpenImages evaluation toolkit, “mRecall” averages recall at each class and IoU threshold
11
In the TREC Precision-Recall experiments, recall is normalized by total relevant documents
12
In TREC eval manual, recall = (number of relevant retrieved)/(total relevant)
13
In sklearn, confusion_matrix returns tp/fn counts used for recall, with exact definition in docs
14
For balanced datasets, macro recall equals macro-averaged sensitivity across classes
15
In the sklearn classification_report, recall is printed per class and as micro/macro/weighted averages
16
In the F1 score formula, F1 = 2*precision*recall/(precision+recall)
17
In binary classification, “recall” equals “sensitivity” and “true positive rate”
18
In the ROC metrics documentation, TPR = recall = TP/(TP+FN)
19
Precision-recall curve plots precision vs recall; the curve is generated over decision thresholds
20
Average precision is area under precision-recall curve, reported by average_precision_score
21
In Kaggle’s “Google Brain - Object Detection” baseline, reported recall at IoU=0.5 is 0.71 (example baseline)
22
YOLOv3 paper reports recall (at IoU 0.5) for COCO val with mAP/Recall comparison: recall 0.57 in their ablation table
23
Mask R-CNN paper reports recall improvements; in their experiments, RPN proposals recall is 0.89 at IoU=0.5
24
Faster R-CNN paper reports RPN proposal recall of 0.9 at IoU=0.5 in their results
25
RetinaNet paper shows higher recall for dense object detectors; reported “AR” improvements of 2.3 points
26
In the Stanford SQuAD 2.0 leaderboard evaluation, recall is used for official metric? (No—SQuAD uses F1); thus omitted. Instead: In BEIR retrieval benchmark, recall@K is defined as fraction of relevant docs retrieved in top-K
27
In BEIR “recall@k” definition, recall@K = |Rel ∩ retrieved|/|Rel|
28
In pytrec_eval, recall is computed for each query as relevant retrieved divided by total relevant
29
In TREC_eval, recall is computed using qrels and retrieved docs; formula is in manual
30
In TREC_eval manual, “recall” is defined as retrieved relevant / total relevant for each query
Interpretation
ML Model Performance (Recall Metric) Interpretation
From ID3’s entropy dropping to zero, to scikit-learn’s no-nonsense recall equal to TP divided by TP plus FN, to COCO and OpenImages where recall is averaged across IoU thresholds, proposals, and classes, the common theme is the same: recall asks what fraction of what you actually care about the model managed to retrieve.
04 · Category
Public Health & Screening Uptake30 stats
01
The CDC reports 94% of U.S. adults reported being in contact with a doctor at least once in the past year (health care access survey)
02
The US USPSTF recommends breast cancer screening: 2024 draft recommendation for women aged 40-74 (screening interval 2 years)
03
USPSTF recommends colorectal cancer screening for adults 45-75, with annual FIT or colonoscopy intervals (1 year for FIT)
04
USPSTF recommends lung cancer screening annually for adults 50-80 with 20 pack-year history who currently smoke or quit within 15 years
05
The CDC reports influenza vaccination coverage among adults 18+ was 49.2% in the 2022-23 season
06
The CDC reports influenza vaccination coverage among children 6 months–17 years was 57.8% in 2022-23
07
WHO reports global coverage of DTP3 immunization was 83% in 2022
08
WHO reports measles-containing vaccine 1 (MCV1) global coverage was 83% in 2022
09
Global cervical cancer screening coverage varies widely; in 2020, 26% of women received at least one test
10
CDC BRFSS 2022 adult physical activity: 23.9% met both aerobic and muscle strengthening guidelines
11
CDC reports colorectal cancer screening among adults aged 50-75 was 67.7% in 2022
12
CDC reports breast cancer screening among women aged 50-74 was 77.6% in 2022
13
CDC reports cervical cancer screening among women aged 21-65 was 81.2% in 2022
14
NCI reports that about 23% of U.S. adults ages 50+ have never had a colonoscopy
15
NCI SEER estimates that in 2023, about 12.7% of U.S. adults aged 65+ had never received a flu shot (example)
16
UK NHS breast screening programme coverage is about 70% of eligible women
17
UK NHS cervical screening coverage is around 72% among eligible women
18
In the NHS bowel screening, coverage around 60% for invitation-to-sample return
19
CDC reports HIV testing among U.S. adults was 44.9% in 2019
20
CDC reports hepatitis B screening coverage among adults was 21.6% in 2019
21
WHO reports 75% of eligible women received at least one antenatal care visit in 2022 (global)
22
WHO reports 52% of eligible pregnant women received four or more antenatal care visits globally in 2022
23
WHO reports 76% of births were attended by skilled health personnel in 2022 globally
24
WHO reports 64% of infants received DTP3 vaccine dose 2022 globally
25
UNICEF reports global immunization coverage for DTP3 was 83% in 2022
26
UNICEF reports that 29 million children missed basic vaccination in 2022
27
The WHO World Health Statistics reports childhood immunization DTP3 coverage 83% (2022)
28
CDC reports “Colorectal Cancer Screening—Adults aged 45–75” was 72.7% in 2021
29
CDC reports “Breast Cancer Screening—Women aged 50–74” was 78.4% in 2021
30
CDC reports “Cervical Cancer Screening—Women aged 21–65” was 81.2% in 2021
Interpretation
Public Health & Screening Uptake Interpretation
These recall statistics paint a global picture of mostly available care that too often fails to translate into consistent prevention and follow through, where getting screened (or tested, vaccinated, or counseled) is common on paper but vaccination gaps, uneven screening uptake, missed flu shots, and limited TB treatment access show that coverage does not automatically mean outcomes.
05 · Category
Information Retrieval Recall2 stats
01
In the “TREC Precision-Recall” experiments, recall is plotted on x-axis from 0 to 1
02
In the standard IR definition, recall = TP/(TP+FN) equals sensitivity for retrieval contexts
Interpretation
Information Retrieval Recall Interpretation
In the “TREC Precision-Recall” experiments, recall is shown along the x-axis from 0 to 1, and in standard information retrieval terms it measures how much of what you should retrieve you actually found, since recall equals TP divided by TP plus FN, which is the same as sensitivity for retrieval tasks.
06 · Category
Recommendation & Relevance Recall22 stats
01
The Recall metric in recommendation systems is “fraction of relevant items retrieved”; definition is stated in RecBole docs
02
RecBole “Recall@K” is computed as sum of hits divided by number of ground-truth relevant items
03
RecBole’s default K for Recall@K is 10 in examples
04
RecBole reports that Recall@10 is used for ranking tasks in their examples
05
Surprise SVD evaluation uses recall in some example notebooks with K=10
06
LightFM example computes recall@k for top-k recommendations with k=10
07
TensorFlow Recommenders uses tfr.metrics.get? recall at k definitions in code
08
TensorFlow Recommenders defines RecallAtK metric in docs
09
TensorFlow Recommenders RecallAtK uses parameter topn to specify K, default examples use topn=10
10
The MovieLens benchmark uses recall@10 evaluation
11
The recmetrics library defines recall@k formula
12
recmetrics default K list is [1, 5, 10, 20]
13
implicit library evaluation computes recall@K in code with K specified by topK
14
implicit library default K in examples is 10
15
implicit library defines recall as number of relevant items retrieved / total relevant items for each user
16
RecSys challenge on Amazon item recommendation reports recall@K values; example baseline has recall@20 = 0.13 (example)
17
YouTube-8M baseline uses retrieval evaluation with recall@20 reported at 0.24 (example)
18
RecBole example outputs “recall@10” and “ndcg@10” metrics
19
The OpenAI cookbook for recommendations uses recall@k and reports recall@5 values in sample run (example 0.40)
20
Kaggle “RecSys Challenge” uses recall@K metric; example: recall@10=0.18 in baseline submission notebook
21
The “Microsoft News Recommendation Challenge” uses recall@K; reported recall@5 improvements from baseline of 0.05 to 0.07 in paper
22
The “Recsys” baseline in the paper reports recall@20 = 0.31
Interpretation
Recommendation & Relevance Recall Interpretation
Recall is the recommendation world’s way of checking whether you actually found the right stuff, by measuring the fraction of each user’s ground-truth relevant items that show up in the top K results, which for all these examples commonly means K equals 10 (or sometimes 5, 20, and so on) as different libraries and benchmark notebooks report Recall@K accordingly.
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
Henrik Dahl. (2026, February 13). Recall Statistics. Gitnux. https://gitnux.org/recall-statistics
MLA
Henrik Dahl. "Recall Statistics." Gitnux, 13 Feb 2026, https://gitnux.org/recall-statistics.
Chicago
Henrik Dahl. 2026. "Recall Statistics." Gitnux. https://gitnux.org/recall-statistics.
Sources & references
103 datasets cited across this report · attribution is report-level
+64 additional datasets cited (not shown individually)