CoChat
Sign Up
Cited by 17Electronic Health Records SystemsMedical Coding and Health InformationNursing Diagnosis and Documentation

Structured and unstructured social risk factor documentation in the electronic health record underestimates patients’ self-reported risks

Open access
Iott, Bradley E, Rivas, Samantha, Gottlieb, Laura M, Adler-Milstein, Julia, Pantell, Matthew S·2024·Journal of the American Medical Informatics Association
DOI 10.1093/jamia/ocad261PMC PMC10873825
Open this paper in CoChat — free accountOpen PDF

Abstract

Objectives: National attention has focused on increasing clinicians’ responsiveness to the social determinants of health, for example, food security. A key step toward designing responsive interventions includes ensuring that information about patients’ social circumstances is captured in the electronic health record (EHR). While prior work has assessed levels of EHR “social risk” documentation, the extent to which documentation represents the true prevalence of social risk is unknown. While no gold standard exists to definitively characterize social risks in clinical populations, here we used the best available proxy: social risks reported by patient survey. Materials and Methods: We compared survey results to respondents’ EHR social risk documentation (clinical free-text notes and International Statistical Classification of Diseases and Related Health Problems [ICD-10] codes). Results: Surveys indicated much higher rates of social risk (8.2%-40.9%) than found in structured (0%-2.0%) or unstructured (0%-0.2%) documentation. Discussion: Ideally, new care standards that include incentives to screen for social risk will increase the use of documentation tools and clinical teams’ awareness of and interventions related to social adversity, while balancing potential screening and documentation burden on clinicians and patients. Conclusion: EHR documentation of social risk factors currently underestimates their prevalence.

AI snapshot

Written from the paper’s full text.

TL;DR

In a pediatric urgent-care sample, patient survey–reported social risks were far more prevalent than social risk documentation in the electronic health record. Structured and unstructured EHR documentation underestimated patients’ self-reported social risks.

Key findings

  • Among 597 caregivers, 85.6% endorsed at least 1 social risk on survey, with an average of 2.5 social risks reported per survey; survey-derived prevalence ranged from 8.2% (job conditions) to 40.9% (housing insecurity).
  • EHR documentation was much lower: ICD-10 Z-code documentation ranged from 0% to 0.2%, free-text note documentation ranged from 0% to 2.0%, and combined structured or unstructured documentation ranged from 0% to 2.0%.
  • Of patients who endorsed at least 1 social risk, 5.5% also had either a Z code or free-text note documenting that risk during the enrollment visit; no patients had concordant documentation in both Z codes and notes.
  • In sensitivity analyses including prior encounters, 15.4% of patients had at least 1 instance of structured or unstructured social risk documentation.

Methods

  • Secondary analysis of data from a completed randomized clinical trial based in a pediatric urgent care center at an urban safety net clinic; the sample included 597 caregivers with no missing data for all included variables.
  • Caregivers completed an 18-item survey of social risk factors; the study compared survey results with EHR documentation of the same 8 social risks using structured ICD-10 Z codes from the problem list at enrollment or in the following 2 days, and unstructured free-text clinical narrative notes associated with the enrollment encounter.
  • A medical student extracted data from 597 notes, with study team reconciliation of ambiguous documentation; analyses tabulated sample demographics and frequencies of survey-endorsed risks versus documentation, with sensitivity analyses including prior Z codes and prior social risk mentions, conducted in Stata 17.0.

Limitations

  • The paper states that Z codes often represent broad social risk concepts rather than specific survey domains, so structured data constructs do not perfectly align with survey questions; generalizability to other pediatric clinics or clinical settings is unclear; clinician documentation behavior may have been influenced by the trial; and the sample was mostly Hispanic and female, which may limit generalizability.
  • The text also makes evident that this was a single urgent care safety-net clinic, and the authors note that survey data may themselves underestimate true social risk prevalence, suggesting the observed EHR documentation deficits may be even greater.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/jamia/ocad261

Cites

31 of 31 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

10 of 25 cited sources checked · 4 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    The literature on social risk factor documentation is hampered by a lack of data about the true prevalence of social risks in the clinical populations being studied.

    “But this literature is hampered by a lack of data about the true prevalence of social risks in the clinical populations being studied.[6][7][8][9]”
    Quote in the paperNot a substantial claim2 of 4 sources checked
    • cites [6] Harle, 2023, Accuracy of electronic health record food insecurity, housing instability, and financial strain screening in adult primary care, JAMA source
    • cites [7] Vale, 2022, Discuss and remember: clinician strategies for integrating social determinants of health in patient records and care, Soc Sci Med source
    • cites [8] Kepper, 2023, The adoption of social determinants of health documentation in clinical settings source

      The Discussion (page not available) says the study could not distinguish patients without social needs from those whose needs were undocumented or recorded outside Z-codes. That supports uncertainty about the true prevalence in this study’s clinical population, but the paper does not establish that the literature generally lacks such data. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites [9] Cottrell, 2018, Using health information technology to bring social determinants of health into primary care: a conceptual framework to guide research, J Health Care Poor Underserved source

      The Abstract (page unavailable) says evidence is limited on how to implement social-determinants screening in primary care. The OCHIN research-portfolio section (page unavailable) discusses collecting patient-reported data, but neither passage says that documentation research is hampered by a lack of data on the true prevalence of social risks in the clinical populations studied. Searches for prevalence and that specific data gap found no passage addressing it. (LLM, conf=0.72)

      Experimental — accuracy not yet measured
  • Findingown data

    Rates of concordant documentation with both Z codes and clinical free-text notes were low in our sample (0%-0.5%).

    “Rates of concordant documentation with both Z codes and clinical free-text notes were low in our sample (0%-0.5%),which may be driven by the burden of documenting the same social risk information in 2 different places in the EHR-“double documenting.””
    Quote in the paperSubstantial claim
  • Findingown data

    The survey-derived prevalence of individual social risks ranged from 8.2% (job conditions) to 40.9% (housing insecurity).

    “The survey-derived prevalence of individual social risks ranged from 8.2% (job conditions) to 40.9% (housing insecurity, Figure 1).”
    Quote in the paperSubstantial claim
  • Findingown data

    In this study, for most included social risk factors, structured SDH documentation was more common than unstructured documentation.

    “In this study, for most included social risk factors, we observe structured SDH documentation to be more common than unstructured documentation.”
    Quote in the paperNot a substantial claim
  • Findingown data

    Surveys indicated much higher rates of social risk (8.2%-40.9%) than found in structured (0%-2.0%) or unstructured (0%-0.2%) documentation.

    “Surveys indicated much higher rates of social risk (8.2%-40.9%)than found in structured (0%-2.0%)or unstructured (0%-0.2%)documentation.”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim. 17
  • May combine two claims. 11

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Clinicians may not be aware of structured SDH documentation tools, and prior work has shown that free-text notes were more commonly used for SDH documentation.

    “Alternatively, clinicians may not be aware of structured SDH documentation tools, 18 and prior work has shown that free-text notes were more commonly used for SDH documentation. 5”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [18] Iott, 2022, Physician awareness of social determinants of health documentation capability in the electronic health record, J Am Med Inform Assoc source
    • cites [5] Navathe, 2018, Hospital readmission and social risk factors identified from physician notes, Health Serv Res source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Prior work investigating social risk factor documentation using ICD-10 codes has suggested that these codes are underutilized, and one study found that ICD codes were used less frequently to document social risk factors when compared to note narratives (unstructured data).

    “Prior work investigating social risk factor documentation using International Statistical Classification of Diseases and Related Health Problems (ICD-10) codes, specifically, has suggested that these codes are underutilized, 3,4 and 1 study found that ICD codes were used less frequently to document social risk factors when compared to note narratives (unstructured data).5”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites [3] Molina, 2023, Social risk factor documentation in emergency departments, Ann Emerg Med source

      Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [4] Pantell, 2020, Effects of in-person navigation to address family social needs on child health care utilization: a randomized clinical trial, JAMA Netw Open source

      Split into 2 atoms — 0 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 2 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [5] Navathe, 2018, Hospital readmission and social risk factors identified from physician notes, Health Serv Res source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The social risk survey instrument used in this study was sourced from a previous social risk factor screening trial.

    “The survey instrument was sourced from a previous social risk factor screening trial and is shown in Appen- dix SA. 11,12”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [11] Gottlieb, 2016, Effects of social needs screening and in-person service navigation on child health: a randomized clinical trial, JAMA Pediatr source
    • cites [12] Gottlieb, 2014, A randomized trial on screening for social determinants of health: the iScreen study, Pediatrics source

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Premisecitation

    Clinicians who lack resources to assist with patients' social risks might choose not to document them, even if they were endorsed by patients during clinical encounters.

    “Clinicians who lack resources to assist with patients' social risks might choose not to document them, even if they were endorsed by patients during clinical encounters.17”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [17] Pantell, 2019, Practice capacity to address patients’ social needs and physician satisfaction and perceived quality of care, Ann Fam Med source

      The Methods describe physicians’ perceived capacity to manage patients’ social needs, and the Discussion relates that capacity to job satisfaction and perceived quality of care. Neither section addresses whether clinicians omit documentation of social risks that patients disclosed; searches for documentation and screening found no such passage. (LLM, conf=0.75)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    There is increasing evidence of the importance of providing social care even in acute emergency settings, suggesting opportunity for social care interventions in urgent care clinics.

    “However, there is increasing evidence of the importance of providing social care even in acute emergency settings, suggesting opportunity for social care interventions in urgent care clinics.38,39”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [38] Anderson, 2016, Addressing social determinants of health from the emergency department through social emergency medicine, West J Emerg Med source
    • cites [39] Walter, 2021, Emergency department-based interventions affecting social determinants of health in the United States: a scoping review, Acad Emerg Med source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Other state and federal policies encourage both SDH screening and the use of structured documentation tools, which may provide computable forms of SDH data for clinical decision support and community resource referral tools.

    “Other state and federal policies encourage both SDH screening and the use of structured documentation tools, which may in turn provide computable forms of SDH data that could be used in clinical decision support algorithms and community resource referral tools.1,35,36”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites [1] National Academies of Sciences, Medicine, 2020

      Not checked in the free run.

    • cites [35] Alderwick, 2019, Meanings and misunderstandings: a social determinants of health Lexicon for health care systems, Milbank Q source

      The Conclusion says federal and state initiatives have strengthened health care systems’ role in identifying and addressing patients’ social needs. The Social Needs-Informed Care and Social Needs-Targeted Care section also describes using screening data to inform clinical decisions and linking patients with social supports. Neither passage says these policies encourage structured documentation tools or that such tools produce computable data for decision-support algorithms and referral tools. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites [36] Cartier

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Findingown data

    We observed no documentation about job conditions in Z codes or free-text notes, though 8.2% of our sample endorsed concerns about job conditions.

    “We observed no documentation about job conditions in Z codes or free-text notes, though 8.2% of our sample endorsed concerns about job conditions.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    85.6% of respondents endorsed at least 1 social risk on the survey; an average of 2.5 social risks were reported per survey.

    “85.6% of respondents endorsed at least 1 social risk on the survey; an average of 2.5 social risks were reported per survey.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given

    Our sample includes 597 caregivers with no missing data for all included variables.

    “Our sample includes 597 caregivers with no missing data for all included variables.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Structured documentation reflected usual medical care and physicians' use of ICD-10 Z codes was not a component of the trial interventions.

    “Structured documentation reflected usual medical care and physicians' use of ICD-10 Z codes was not a component of the trial interventions, as a separate researcher administered the survey and provided resource referrals.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data

    Definitions for each social risk factor were modeled from the patient survey questions and inclusion/exclusion decisions were consensus-based.

    “We modeled definitions for each social risk factor from the patient survey questions (Appendix SA) and decisions for inclusion or exclusion were consensus-based.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data

    Data extraction from 597 notes was performed by a medical student, and members of the study team met to reconcile extracted data, review ambiguous documentation, and make inclusion decisions.

    “A medical student (S.R.) extracted data from 597 notes and members of the study team (B.E.I., S.R., and M.S.P.) met to reconcile extracted data, review instances of ambiguous documentation, and make decisions about inclusion criteria for each type of social risk factor.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    This study was approved by the University of California San Francisco Institutional Review Board.

    “This study was approved by the University of California San Francisco Institutional Review Board.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Unstructured documentation reflected usual medical care and physicians' use of clinical free-text notes was not a component of the trial interventions.

    “Similarly, unstructured documentation reflected usual medical care and physicians' use of clinical free-text notes was not a component of the trial interventions, as a separate researcher administered all study activities.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Decisions about inclusion criteria were kept in an evolving document shared with the study team to guide ongoing data extraction.

    “Decisions about inclusion criteria were kept in an evolving document shared with the study team to guide ongoing data extraction.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    New social care quality measures will require reporting on the prevalence of screening for social risks.

    “New social care quality measures will require reporting on the prevalence of screening for social risks.[27][28][29][30][31][32][33][34]”
    Quote in the paperSubstantial claim0 of 8 sources checked
    • cites [27] Hayes, 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [28] National Committee for Quality Assurance, 2021

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [29] PatientEngagementHIT, 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [30] Nuzum, 2021 source
    • cites [31] Centers for Medicare & Medicaid Services, 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [32] 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [33] The Joint Commission, 2021

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [34] The Joint Commission

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Most of our sample of caregivers identified as Hispanic and female, and this sample may not generalize to certain patient populations.

    “Finally, most of our sample of caregivers identified as Hispanic and female, and this sample may not generalize to certain patient populations.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    We cannot be sure that clinician documentation behavior was not influenced by the trial.

    “Additionally, while trial intervention arms were carried out by separate researchers, we cannot be sure that clinician documentation behavior was not influenced by the trial.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Study design and conductnone given

    It is not clear the extent to which our findings generalize to other pediatric clinics or clinical settings.

    “Additionally, it is not clear the extent to which our findings generalize to other pediatric clinics or clinical settings.”
    Quote in the paperNot a substantial claim
19 of 19

  • The paper relies on this without support.

    No gold standard exists to definitively characterize social risks in clinical populations.

    “While no gold standard exists to definitively characterize social risks in clinical populations”

  • The paper relies on this without support.

    Patient survey is the best available proxy for the true prevalence of social risks.

    “here we used the best available proxy: social risks reported by patient survey”

  • The paper relies on this without support.

    Survey data may underestimate the true prevalence of social risks.

    “Since survey data also may underestimate the true prevalence of social risks”

  • The paper relies on this without support.

    The extent to which EHR documentation represents the true prevalence of social risk is unknown.

    “the extent to which documentation represents the true prevalence of social risk is unknown”

  • The paper relies on this without support.

    This is the first study to directly compare rates of both structured and unstructured social risk factor EHR documentation with patient-endorsed social risks collected by survey.

    “This is the first study of which we are aware to directly compare rates of both structured and unstructured social risk factor EHR documentation with patient-endorsed social risks collected by survey.”

  • The paper relies on this without support.

    The pattern of underdocumentation of social risks is similar in other settings.

    “we believe that the pattern of underdocumentation of social risks is similar in other settings”
6 of 6

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Abstract reverses structured and unstructured documentation ranges

    Minor

    The abstract assigns the 0%-2.0% range to structured documentation and the 0%-0.2% range to unstructured documentation. The Results assign 0%-0.2% to ICD-10 Z codes and 0%-2.0% to free-text notes.

  • Discussion reverses which documentation type was more common

    Minor

    The Discussion says structured documentation was more common for most social risks, whereas the Results say free-text documentation was more common for most social risks.

2 of 2

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. National Academies of Sciences, Medicine, 2020, Integrating Social Care into the Delivery of Health Care: Moving Upstream to Improve the Nation’s Health

Cited by

Sorted by citations

Related

Authors
Iott, B. E., Rivas, S., Gottlieb, L. M., Adler-Milstein, J., & Pantell, M. S.
Affiliations
  • University of California, San Francisco (US)
  • Center for Clinical Informatics and Improvement Research, University of California, San Francisco , San Francisco, CA, United States
Corresponding author
Iott, Bradley E
ORCID
Venue
Journal of the American Medical Informatics Association
Publisher
Oxford University Press (OUP)
Published
2024-01-12
Volume
31
Issue
3
Pages
714-719
ISSN
  • 1067-5027 (ISSN-L)
  • 1527-974X
DOI
10.1093/jamia/ocad261
PMC
PMC10873825
OpenAlex
W4390848828
Semantic Scholar
8ceaed722bbae3d721edee759d381879ba4a53c2
Cited by
17
References
28
Citations by year
2026: 10 · 2025: 3 · 2024: 4
FWCI
5.04
Type
journal article
Language
en
Open access
Licence
academic.oup.com/pages/standard-publication-reuse-rights (vor)
Locations
  • Journal of the American Medical Informatics Association (publishedVersion)
  • PubMed (publishedVersion)
  • eScholarship (California Digital Library) (submittedVersion)
  • PubMed Central (submittedVersion)
Topics
  • Health Sciences › Health Professions › Health Information Management › Electronic Health Records Systems
  • Health Sciences › Health Professions › Health Information Management › Medical Coding and Health Information
  • Health Sciences › Nursing › Issues, ethics and legal aspects › Nursing Diagnosis and Documentation
Keywords
social risk factors, social determinants of health, electronic health records, food insecurity, patient self-report, ICD-10 codes, documentation burden
MeSH terms
Documentation, Humans, Risk Factors, Prevalence, Electronic Health Records, Self Report

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-27.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-27.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review