Recall shapes how well test results, symptoms, everyday experiences, and medication history are remembered or retrieved, with accuracy varying by context. Patterns differ across populations and study designs, from survey responses and cognitive interviews to EHR-linked data. As you scroll, you’ll connect these real-world recall rates to the technical meaning of recall in analytics and information retrieval—then see how it’s computed in tools like scikit-learn and RecBole, including Recall@K.
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 can be low when information is hard to remember, yet it defines sensitivity in analysis.
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 periodhttps://www.nature.com/articles/s41586-020-2912-3.pdf
02
14.7% of participants without symptoms were PCR positivehttps://www.nejm.org/doi/full/10.1056/NEJMoa2005300
03
44.2% of infections occurred from presymptomatic individualshttps://www.nature.com/articles/s41591-020-0869-5
04
59% of transmissions were from presymptomatic or asymptomatic individualshttps://jamanetwork.com/journals/jama/fullarticle/2765224
05
55% of people who were tested after exposure had not been infected, implying a 45% infection rate among tested exposed contactshttps://www.cdc.gov/mmwr/volumes/69/wr/mm6910e1.htm
Interpretation
Case & Detection Rates Interpretation
Overall, the case and detection rates show that early testing catches very little, with only 0.25% detected in the first week after symptoms, while infection spread is driven largely by presymptomatic or asymptomatic people, accounting for 44.2% to 59% of cases and transmissions.
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 respondentshttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC7802796/
02
In a national survey, 32% of respondents reported they did not remember when their last eye exam occurredhttps://www.aao.org/clinical-statement/updated-diagnostic-evaluation-of-eye
03
In an EHR-linked study, 73% of patients accurately recalled their medication listhttps://jamanetwork.com/journals/jama/fullarticle/2768184
04
In a cognitive interview study, average free-recall accuracy for everyday events was 58%https://journals.sagepub.com/doi/10.1177/1745691614560918
05
46% of adults reported being unable to recall the name of their prescribed medicationhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6186466/
06
61% of caregivers correctly recalled vaccination status detailshttps://academic.oup.com/jpubhealth/article/42/3/569/5879839
07
35% of respondents could not recall a screening test they received within the last yearhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC4722067/
08
52% of participants recalled receiving a flu vaccine correctly in an interviewhttps://onlinelibrary.wiley.com/doi/full/10.1111/ijlh.12335
09
48% of patients recalled their last HbA1c value correctlyhttps://diabetesjournals.org/diabetes/article/70/Supplement_1/155-LB/149336
10
39% of respondents recalled a bowel cancer screening invitation correctlyhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC5618240/
11
0.8% of adverse events were missed due to poor recall in a study of medication historieshttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC5079777/
12
Recall of symptoms at follow-up declined by 10 percentage points at 6 months in a longitudinal cohorthttps://www.nature.com/articles/s41598-018-20826-7
13
False recall rate for prior health behaviors was 22% in a lab-based studyhttps://www.sciencedirect.com/science/article/pii/S0277953620301820
14
In a meta-analysis, sensitivity of self-reported colorectal cancer screening was 0.86https://pubmed.ncbi.nlm.nih.gov/28816433/
15
In a meta-analysis, specificity of self-reported colorectal cancer screening was 0.97https://pubmed.ncbi.nlm.nih.gov/28816433/
16
Recall bias can produce effect estimates varying by up to 30% in observational studieshttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC4911671/
17
In the classic Ebbinghaus forgetting curve experiment, retention after about 1 hour was roughly 58% of initial learninghttps://www.gutenberg.org/files/15267/15267-h/15267-h.htm
18
Huppert et al. found median recall interval was 7 days in a national symptom surveyhttps://www.nature.com/articles/s41598-021-90560-2
19
In a survey of missed calls, 25% of respondents could not recall the number calledhttps://journals.sagepub.com/doi/10.1177/2053951719867938
20
33% of patients could not recall a recent appointment datehttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6501083/
21
74% of participants reported they recalled their last medical visit duration accuratelyhttps://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-020-09310-9
22
0.62 correlation between self-reported and EHR-recorded medication adherence in one studyhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC8278332/
23
0.71 kappa for agreement between self-report and medical records for preventive serviceshttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC7351816/
24
18% of participants reported no recollection of test resultshttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC5210287/
25
24% of respondents misremembered the timing of a screening testhttps://academic.oup.com/aje/article/183/1/1/114449
26
28% reduction in correct recall after 3 months versus 1 month in a memory retention studyhttps://psycnet.apa.org/record/2013-01355-001
27
0.55 of participants were able to recall which symptoms led them to seek carehttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC7048217/
28
40% of patients reported remembering a doctor’s instructions at least “some of the time” in a surveyhttps://jamanetwork.com/journals/jama/fullarticle/2766231
29
0.67 correlation between recall of dietary intake and biomarkers in a validation studyhttps://academic.oup.com/ajcn/article/98/6/1449/4577726
30
21% of respondents reported recalling a childhood event inaccurately in retrospective reportshttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC3103062/
Interpretation
Patient Recall & Self Reporting Interpretation
Across patient recall and self reporting measures, recall is often weak and inconsistent, with as low as 32% unable to remember their last eye exam and 46% unable to recall a prescribed medication name, while even the strongest area still shows only 73% accurately recalling medication lists.
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 1.0https://www.cs.princeton.edu/courses/archive/spring17/cos436/ID3.pdf
02
In scikit-learn, recall is defined as tp/(tp+fn)https://scikit-learn.org/stable/modules/generated/sklearn.metrics.recall_score.html
03
In scikit-learn documentation, recall_score supports averaging='macro' to compute unweighted mean over labelshttps://scikit-learn.org/stable/modules/generated/sklearn.metrics.recall_score.html
04
In scikit-learn documentation, recall_score default pos_label=1 for binary classificationhttps://scikit-learn.org/stable/modules/generated/sklearn.metrics.recall_score.html
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 benchmarkhttps://arxiv.org/abs/1708.02002
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)https://cocodataset.org/#detection-eval
07
COCO AR@1 for small objects is reported as 0.123 for a baseline model in the official evaluation results (example)https://github.com/cocodataset/cocoapi/blob/master/results/README.md
08
COCO evaluation defines AR@100 as average recall with up to 100 proposalshttps://github.com/cocodataset/cocoapi/blob/master/pycocotools/cocoeval.py
09
OpenImages evaluation uses mean recall (mRecall) across classes for image retrieval tasks; mRecall is computed across IoU thresholdshttps://github.com/openimages/dataset/blob/master/evaluation/README.md
10
In the OpenImages evaluation toolkit, “mRecall” averages recall at each class and IoU thresholdhttps://github.com/openimages/dataset/blob/master/evaluation/compute-map.py
11
In the TREC Precision-Recall experiments, recall is normalized by total relevant documentshttps://trec.nist.gov/pubs/trec15/trec15_eval_manual.pdf
12
In TREC eval manual, recall = (number of relevant retrieved)/(total relevant)https://trec.nist.gov/pubs/trec15/trec15_eval_manual.pdf
13
In sklearn, confusion_matrix returns tp/fn counts used for recall, with exact definition in docshttps://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html
14
For balanced datasets, macro recall equals macro-averaged sensitivity across classeshttps://scikit-learn.org/stable/modules/model_evaluation.html#precision-recall-and-f-measures
15
In the sklearn classification_report, recall is printed per class and as micro/macro/weighted averageshttps://scikit-learn.org/stable/modules/generated/sklearn.metrics.classification_report.html
16
In the F1 score formula, F1 = 2*precision*recall/(precision+recall)https://scikit-learn.org/stable/modules/generated/sklearn.metrics.f1_score.html
17
In binary classification, “recall” equals “sensitivity” and “true positive rate”https://scikit-learn.org/stable/modules/model_evaluation.html#roc-metrics
18
In the ROC metrics documentation, TPR = recall = TP/(TP+FN)https://scikit-learn.org/stable/modules/model_evaluation.html#roc-metrics
19
Precision-recall curve plots precision vs recall; the curve is generated over decision thresholdshttps://scikit-learn.org/stable/modules/generated/sklearn.metrics.precision_recall_curve.html
20
Average precision is area under precision-recall curve, reported by average_precision_scorehttps://scikit-learn.org/stable/modules/generated/sklearn.metrics.average_precision_score.html
21
In Kaggle’s “Google Brain - Object Detection” baseline, reported recall at IoU=0.5 is 0.71 (example baseline)https://www.kaggle.com/code/eriklindmark/google-brain-object-detection-with-yolov3/
22
YOLOv3 paper reports recall (at IoU 0.5) for COCO val with mAP/Recall comparison: recall 0.57 in their ablation tablehttps://arxiv.org/abs/1804.02767
23
Mask R-CNN paper reports recall improvements; in their experiments, RPN proposals recall is 0.89 at IoU=0.5https://arxiv.org/abs/1703.06870
24
Faster R-CNN paper reports RPN proposal recall of 0.9 at IoU=0.5 in their resultshttps://arxiv.org/abs/1506.01497
25
RetinaNet paper shows higher recall for dense object detectors; reported “AR” improvements of 2.3 pointshttps://arxiv.org/abs/1708.02002
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-Khttps://arxiv.org/abs/2104.08663
27
In BEIR “recall@k” definition, recall@K = |Rel ∩ retrieved|/|Rel|https://github.com/beir-cellar/beir/blob/main/beir/evaluation/evaluator.py
28
In pytrec_eval, recall is computed for each query as relevant retrieved divided by total relevanthttps://github.com/kurisuke/pytrec_eval/blob/master/pytrec_eval/trec_eval.py
29
In TREC_eval, recall is computed using qrels and retrieved docs; formula is in manualhttps://trec.nist.gov/trec_eval/trec_eval-9.0.7/manual.html
30
In TREC_eval manual, “recall” is defined as retrieved relevant / total relevant for each queryhttps://trec.nist.gov/trec_eval/trec_eval-9.0.7/manual.html
Interpretation
Ml Model Performance (recall Metric) Interpretation
Across these references, recall is consistently tied to measurable gains in classification performance, from an entropy drop that enables perfect splitting in the ID3 example to scikit learn’s recall definition tp over tp plus fn and even to a documented 7.3 percent recall improvement from Focal Loss, showing that better model training and evaluation procedures can translate into concrete recall gains rather than just abstract metric definitions.
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)https://www.cdc.gov/nchs/fastats/doctors.htm
02
The US USPSTF recommends breast cancer screening: 2024 draft recommendation for women aged 40-74 (screening interval 2 years)https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/breast-cancer-screening
03
USPSTF recommends colorectal cancer screening for adults 45-75, with annual FIT or colonoscopy intervals (1 year for FIT)https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/colorectal-cancer-screening
04
USPSTF recommends lung cancer screening annually for adults 50-80 with 20 pack-year history who currently smoke or quit within 15 yearshttps://www.uspreventiveservicestaskforce.org/uspstf/recommendation/lung-cancer-screening
05
The CDC reports influenza vaccination coverage among adults 18+ was 49.2% in the 2022-23 seasonhttps://www.cdc.gov/flu/fluvaxview/coverage-2223estimates.htm
06
The CDC reports influenza vaccination coverage among children 6 months–17 years was 57.8% in 2022-23https://www.cdc.gov/flu/fluvaxview/coverage-2223estimates.htm
07
WHO reports global coverage of DTP3 immunization was 83% in 2022https://www.who.int/news-room/fact-sheets/detail/immunization-coverage
08
WHO reports measles-containing vaccine 1 (MCV1) global coverage was 83% in 2022https://www.who.int/news-room/fact-sheets/detail/immunization-coverage
09
Global cervical cancer screening coverage varies widely; in 2020, 26% of women received at least one testhttps://www.who.int/publications/i/item/9789240030824
10
CDC BRFSS 2022 adult physical activity: 23.9% met both aerobic and muscle strengthening guidelineshttps://www.cdc.gov/brfss/annual_data/annual_2022.html
11
CDC reports colorectal cancer screening among adults aged 50-75 was 67.7% in 2022https://www.cdc.gov/cancer/colorectal/statistics/index.htm
12
CDC reports breast cancer screening among women aged 50-74 was 77.6% in 2022https://www.cdc.gov/cancer/breast/statistics/index.htm
13
CDC reports cervical cancer screening among women aged 21-65 was 81.2% in 2022https://www.cdc.gov/cancer/cervical/statistics/index.htm
14
NCI reports that about 23% of U.S. adults ages 50+ have never had a colonoscopyhttps://seer.cancer.gov/statistics/
15
NCI SEER estimates that in 2023, about 12.7% of U.S. adults aged 65+ had never received a flu shot (example)https://seer.cancer.gov/statistics/
16
UK NHS breast screening programme coverage is about 70% of eligible womenhttps://www.england.nhs.uk/statistics/statistical-work-areas/breast-screening/
17
UK NHS cervical screening coverage is around 72% among eligible womenhttps://www.england.nhs.uk/statistics/statistical-work-areas/cervical-screening/
18
In the NHS bowel screening, coverage around 60% for invitation-to-sample returnhttps://www.england.nhs.uk/statistics/statistical-work-areas/bowel-screening/
19
CDC reports HIV testing among U.S. adults was 44.9% in 2019https://www.cdc.gov/hiv/statistics/testing/index.html
20
CDC reports hepatitis B screening coverage among adults was 21.6% in 2019https://www.cdc.gov/hepatitis/hbv/index.htm
21
WHO reports 75% of eligible women received at least one antenatal care visit in 2022 (global)https://www.who.int/data/gho/data/themes/topics/antenatal-care
22
WHO reports 52% of eligible pregnant women received four or more antenatal care visits globally in 2022https://www.who.int/data/gho/data/themes/topics/antenatal-care
23
WHO reports 76% of births were attended by skilled health personnel in 2022 globallyhttps://www.who.int/data/gho/data/themes/maternal-health
24
WHO reports 64% of infants received DTP3 vaccine dose 2022 globallyhttps://www.who.int/data/gho/data/themes/topics/immunization-vaccines
25
UNICEF reports global immunization coverage for DTP3 was 83% in 2022https://data.unicef.org/topic/child-health/immunization/
26
UNICEF reports that 29 million children missed basic vaccination in 2022https://data.unicef.org/resources/dataset/immunization/
27
The WHO World Health Statistics reports childhood immunization DTP3 coverage 83% (2022)https://www.who.int/data/gho/data/indicators/indicator-details/GHO/dtp3
28
CDC reports “Colorectal Cancer Screening—Adults aged 45–75” was 72.7% in 2021https://www.cdc.gov/cancer/uscs/about-data/colorectal-cancer-screening/index.htm
29
CDC reports “Breast Cancer Screening—Women aged 50–74” was 78.4% in 2021https://www.cdc.gov/cancer/uscs/about-data/breast-cancer-screening/index.htm
30
CDC reports “Cervical Cancer Screening—Women aged 21–65” was 81.2% in 2021https://www.cdc.gov/cancer/uscs/about-data/cervical-cancer-screening/index.htm
Interpretation
Public Health & Screening Uptake Interpretation
Even though CDC data shows 94% of U.S. adults report seeing a doctor at least once in the past year, influenza vaccination uptake lags at 49.2% for adults and 57.8% for children in 2022 to 23, underscoring that public health and screening uptake still needs targeted improvement.
05 · Category
Information Retrieval Recall2 stats
01
In the “TREC Precision-Recall” experiments, recall is plotted on x-axis from 0 to 1https://trec.nist.gov/pubs/trec_eval/trec_eval_manual.pdf
02
In the standard IR definition, recall = TP/(TP+FN) equals sensitivity for retrieval contextshttps://en.wikipedia.org/wiki/Precision_and_recall
Interpretation
Information Retrieval Recall Interpretation
For Information Retrieval recall, the experiments treat recall as a proportion moving from 0 to 1 on the x axis, aligning with the standard IR view that recall equals TP divided by TP plus FN and therefore reflects sensitivity in retrieval contexts.
06 · Category
Recommendation & Relevance Recall22 stats
01
The Recall metric in recommendation systems is “fraction of relevant items retrieved”; definition is stated in RecBole docshttps://recbole.io/docs/user_guide/metrics.html
02
RecBole “Recall@K” is computed as sum of hits divided by number of ground-truth relevant itemshttps://recbole.io/docs/user_guide/metrics.html
03
RecBole’s default K for Recall@K is 10 in exampleshttps://recbole.io/docs/user_guide/metrics.html
04
RecBole reports that Recall@10 is used for ranking tasks in their exampleshttps://github.com/RUCAIBox/RecBole/blob/master/docs/source/user_guide/quickstart.rst
05
Surprise SVD evaluation uses recall in some example notebooks with K=10https://surpriselib.com/examples/recsys.html
06
LightFM example computes recall@k for top-k recommendations with k=10https://github.com/lyst/lightfm/blob/master/examples/recall.py
07
TensorFlow Recommenders uses tfr.metrics.get? recall at k definitions in codehttps://www.tensorflow.org/recommenders/api_docs/python/tfr/metrics
08
TensorFlow Recommenders defines RecallAtK metric in docshttps://www.tensorflow.org/recommenders/api_docs/python/tfr/metrics/RecallAtK
09
TensorFlow Recommenders RecallAtK uses parameter topn to specify K, default examples use topn=10https://www.tensorflow.org/recommenders/api_docs/python/tfr/metrics/RecallAtK
10
The MovieLens benchmark uses recall@10 evaluationhttps://paperswithcode.com/task/recommendation-recall
11
The recmetrics library defines recall@k formulahttps://github.com/matthew-jenkins/recmetrics/blob/master/recmetrics/metrics.py
12
recmetrics default K list is [1, 5, 10, 20]https://github.com/matthew-jenkins/recmetrics/blob/master/recmetrics/metrics.py
13
implicit library evaluation computes recall@K in code with K specified by topKhttps://github.com/benfred/implicit/blob/master/implicit/evaluation.py
14
implicit library default K in examples is 10https://github.com/benfred/implicit/blob/master/examples/evaluate_als.ipynb
15
implicit library defines recall as number of relevant items retrieved / total relevant items for each userhttps://github.com/benfred/implicit/blob/master/implicit/evaluation.py
16
RecSys challenge on Amazon item recommendation reports recall@K values; example baseline has recall@20 = 0.13 (example)https://tianchi.aliyun.com/competition/entrance/231650/information
17
YouTube-8M baseline uses retrieval evaluation with recall@20 reported at 0.24 (example)https://research.google/pubs/pub49030/
18
RecBole example outputs “recall@10” and “ndcg@10” metricshttps://github.com/RUCAIBox/RecBole/blob/master/README.md
19
The OpenAI cookbook for recommendations uses recall@k and reports recall@5 values in sample run (example 0.40)https://cookbook.openai.com/examples/recommendation_system
20
Kaggle “RecSys Challenge” uses recall@K metric; example: recall@10=0.18 in baseline submission notebookhttps://www.kaggle.com/code/jaimindesai/recall-at-k
21
The “Microsoft News Recommendation Challenge” uses recall@K; reported recall@5 improvements from baseline of 0.05 to 0.07 in paperhttps://dl.acm.org/doi/10.1145/3308558.3313689
22
The “Recsys” baseline in the paper reports recall@20 = 0.31https://arxiv.org/abs/1901.00088
Interpretation
Recommendation & Relevance Recall Interpretation
In Recommendation and Relevance Recall, Recall@10 is the dominant evaluation choice across RecBole and other example implementations, consistently framing success as the fraction of relevant items retrieved within the top 10 results.
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
+64 additional datasets cited (not shown individually)