Abstract
Objectives Duplicate patient records can increase costs and medical errors. We assessed the association between demographic factors, comorbidities, health care usage, and duplicate electronic health records.
Methods We analyzed the association between duplicate patient records and multiple demographic variables (race, Hispanic ethnicity, sex, and age) as well as the Charlson Comorbidity Index (CCI), number of diagnoses, and number of health care encounters. The study population included 3,018,413 patients seen at a large urban academic medical center with at least one recorded diagnosis. Duplication of patient medical records was determined by using a previously validated enterprise Master Person Index.
Results Unknown or missing demographic data, Black race when compared with White race (odds ratio [OR]: 1.35, p < 0.001), Hispanic compared with non-Hispanic ethnicity (OR: 1.48, p < 0.001), older age (OR: 1.01, p < 0.001), and “Other” sex compared with female sex (OR: 4.71, p < 0.001) were associated with higher odds of having a duplicate record. Comorbidities (CCI, OR: 1.10, p < 0.001) and more encounters with the health care system (OR: 1.01, p < 0.001) were also associated with higher odds of having a duplicate record. In contrast, male sex compared with female sex was associated with lower odds of having a duplicate record (OR: 0.88, p < 0.001).
Conclusion The odds of duplications in medical records were higher in Black, Hispanic, older, nonmale patients with more health care encounters, more comorbidities, and unknown demographic data. Understanding the epidemiology of duplicate records can help guide prevention and mitigation efforts for high-risk populations. Duplicate records can contribute to disparities in health care outcomes for minority populations.
Keywords: medical records systems, computerized; patient safety; health care disparities; race factors; epidemiology; clinical data management; clinical information systems
Background and Significance
Medical errors have been estimated to be the third leading cause of mortality in the United States, resulting in more than 250,000 annual deaths. 1 While the majority of these errors are due to diagnostic or therapeutic failure, 2 3 a new set of unforeseen errors, associated with the expanding integration of electronic health records (EHRs) into clinical workflows, has emerged. 4 With the growing adoption of EHR systems, some errors may be attributable to the duplication of patient medical records. 5
Duplicate patient records, defined as two or more records belonging to a single patient, 6 present a significant challenge for organizations dealing with inaccurate patient information from various systems. While the rate of record duplication within a system can vary, with some studies reporting duplication rates up to 16%, 7 there is evidence that duplicate records have negative effects on both hospital finances and patient outcomes. 8 Clinically, duplicate records have been associated with multiple types of harm to patients, such as missed abnormal laboratory results 9 and blood transfusion errors, 10 necessitating laborious and costly measures to find and fix these record duplications. 11 Record duplication can occur for a variety of reasons, including inefficient search capabilities in EHR systems, 12 untrained staff, 13 and complex emergency department environments. 14 Interestingly, changing patient demographics have also been linked to medical record duplication. 15
Objectives
Currently, the associations between record duplication and patient demographics or health care usage are not well understood. In this study, we used patient data in a clinical data warehouse (CDW) to compare demographic and clinical characteristics of patients with and without duplicate records. Since duplicate records are not likely to occur randomly, we hypothesized that duplicate records will be associated with greater health care utilization due to an increased number of entries into the record associated with a single patient.
Methods
We performed a retrospective analysis of patients from the Allscripts Enterprise EHR at the University of Texas Health Science Center at Houston. This study has been approved by the Committee for the Protection of Human Subjects (the UTHSC-H IRB) under protocol HSC-SBMI-13–0549. Data were accessed via a CDW that contained data from the Allscripts EHR. Identity for the CDW was managed by a locally developed enterprise Master Person Index (eMPI). The eMPI used a deterministic rule-based algorithm optimized to local data to assign a unique identification (UID) to each patient. The eMPI identified potential duplicate records (called “local IDs” or LIDs) based on various identifiers including medical record numbers, phone numbers as well as demographic variables such as first name, last name, and date of birth. Duplicate records (LIDs) were linked to a single UID. The eMPI was validated using a set of 20,000 manually reviewed record pairs. Based on this evaluation, the worst-case error for the eMPI as configured for this project was 2.5%. Worst-case error would occur if all miscategorizations were false negatives (i.e., unrecognized duplicate records, therefore falsely undercounting duplicate records). 16
Patient data were retrieved from the CDW using Structured Query Language (SQL) via Microsoft SQL Server 2016 and processed in R version 4.3.2. As shown in Fig. 1 , patients were included if they had at least one ICD-10-CM billing code between January 1, 2015 and January 1, 2023 in at least one record (LID), but no ICD-9-CM billing codes in Allscripts. Patients with ICD-9-CM codes were removed to maximize the accuracy of Charlson Comorbidity Index (CCI) calculations and minimize coding variability. A patient (UID) with two or more LIDs was considered to have a duplicate record.
Fig. 1.
CONSORT diagram of patients included in study.
Patients meeting the above eligibility criteria were split into two groups: patients with versus without duplicate records. For patients with duplicate records that had inconsistent demographic data across LIDs (e.g., different birth dates or last names), data associated with the first admission record were analyzed. For patients with duplicate records, age was calculated at the time of record duplication (i.e., when the second record was created), and for patients without duplicates, age was calculated on the day of last admission. Patients with Hispanic ethnicity values of “other” and ‘unknown’ were combined because only 1,232 patients had a Hispanic status of “other.”
CCI is based on the number of significant comorbidities and has been validated to predict mortality. 17 In this project, we used CCI as a standardized measure of comorbidity burden. CCI was calculated using the R comorbidity package version 1.0.7. 18 A health care encounter was defined as an outpatient visit, emergency department visit, or inpatient hospital stay. The number of encounters was therefore calculated as the total number of encounters during the study period for that patient. We used the Pearson's chi-squared test to evaluate associations between categorical variables including demographic data (race, Hispanic ethnicity, sex) and patient record duplication. We used the Wilcoxon rank-sum test to evaluate associations between ordinal and continuous variables (age, CCI, number of diagnoses, and number of health care encounters) and duplicate records. Odds ratios (ORs) of record duplication were calculated using a generalized linear model with a binomial distribution. Statistical significance (α) was set at 0.05.
Results
A total of 1,185,752 patients were included in this study ( Table 1 ). For patients whose race was recorded, the most frequent group identified as White (36%). Notably, race could not be assigned for a significant minority of patients (44.6%), including 39% “other” and 5.6% “unknown.” Most patients identified as non-Hispanic (84%). Our sample included more females (54%) than males (46%), with less than 0.1% of patient records having a sex value of “other.” Patients with duplicate EHR records made up 2.9% of the data. Interestingly, of patients with duplicate records, 87.8% had a duplicate record with no ICD-10-CM diagnosis code suggesting that the duplicate record was not associated with a billed clinical encounter in the system.
Table 1. Characteristics of the study population.
| Variable | N = 1,185,752 |
|---|---|
| Race | |
| Asian | 22,686 (1.9%) |
| Black | 195,460 (16%) |
| Native American | 4,862 (0.4%) |
| Other | 466,563 (39%) |
| Unknown | 66,883 (5.6%) |
| White | 429,298 (36%) |
| Hispanic | |
| No | 998,939 (84%) |
| Other/unknown | 45,822 (3.9%) |
| Yes | 140,991 (12%) |
| Sex | |
| Female | 640,068 (54%) |
| Male | 545,421 (46%) |
| Other | 263 (<0.1%) |
| Group | |
| Duplicates | 33,815 (2.9%) |
| Nonduplicates | 1,151,937 (97%) |
Note: n (%).
The relationship between record duplication and patient demographic characteristics or health care utilization was initially assessed via single-variable analyses ( Table 2 ). Unknown race was associated with having at least one duplicate record ( p < 0.001). Hispanic ethnicity was associated with having at least one duplicate record ( p < 0.001). Similar to missing race, patients with duplicate records disproportionately had missing ethnicity values such as “other” or “unknown.”
Table 2. Univariable associations with duplicate records.
| Variable | Duplicates ( N = 33,815) | Nonduplicates ( N = 1,151,937) | p -Value a |
|---|---|---|---|
| Race | <0.001 | ||
| Asian | 450 (1.3%) | 22,236 (1.9%) | |
| Black | 5,013 (15%) | 190,447 (17%) | |
| Native American | 103 (0.3%) | 4,759 (0.4%) | |
| Other | 12,183 (36%) | 454,380 (39%) | |
| Unknown | 6,946 (21%) | 59,937 (5.2%) | |
| White | 9,120 (27%) | 420,178 (36%) | |
| Hispanic | <0.001 | ||
| No | 25,293 (75%) | 973,646 (85%) | |
| Other/unknown | 3,554 (11%) | 42,268 (3.7%) | |
| Yes | 4,968 (15%) | 136,023 (12%) | |
| Sex | <0.001 | ||
| Female | 19,221 (57%) | 620,847 (54%) | |
| Male | 14,554 (43%) | 530,867 (46%) | |
| Other | 40 (0.1%) | 223 (<0.1%) | |
| Age | 43 (21) | 38 (24) | <0.001 |
| Unknown | 711 | 2 | |
| CCI | 0.37 (0.91) | 0.26 (0.75) | <0.001 |
| Number of diagnoses | 25 (79) | 19 (62) | <0.001 |
| Encounters | 5 (7) | 4 (6) | <0.001 |
Note: n (%); mean (standard deviation).
Pearson's chi-squared test; Wilcoxon rank-sum test.
Age was significantly associated with record duplication ( p < 0.001). The average age for patients with duplicates was 43 at the time that the duplicate record was created compared with 38 (at the time of last encounter) for patients without duplicate records. Most patients with unknown ages (i.e., no recorded birth date) had duplicate records, although only 713 patients in our sample had no birth date. Patients with duplicates were, on average, sicker than patients without duplicates; average CCI 0.37 vs. 0.26 ( p < 0.001). Patients with duplicates also had more diagnoses (mean 25 vs. 19, p < 0.001) and more health care encounters (5 vs. 4, p < 0.001).
Table 3 (multivariable analysis) shows the ORs of having duplicate records based on a generalized linear model including race, ethnicity, sex, age, CCI, count of diagnoses, and number of encounters. A Hispanic status value of “other” or “unknown” compared with non-Hispanic was associated with having a duplicate record (OR: 2.57 [95% confidence interval, CI: 2.48–2.67]). Similarly, “unknown” race compared with White (OR: 5.28 [95% CI: 5.11–5.46]) and “other” compared with female sex (OR: 4.71 [95% CI: 3.25–6.66]) were also associated with duplicate records. Black race, but not other race values, was associated with higher odds of having a duplicate (OR: 1.35 [95% CI: 1.30–1.40]) compared with White race. Higher CCI, age, and number of encounters were associated with small but statistically significant increases in OR of having a duplicate. In contrast, the number of diagnoses was not significantly associated with an increased odds of duplication on multivariable analysis; perhaps because the number of diagnoses was strongly associated with other comorbidity measures such as CCI and number of encounters.
Table 3. Multivariable associations with duplicate records.
| Variable | OR | 95% confidence interval | p -Value |
|---|---|---|---|
| Race | |||
| White | |||
| Asian | 0.94 | 0.85, 1.03 | 0.2 |
| Black | 1.35 | 1.30, 1.40 | <0.001 |
| Native American | 1.09 | 0.88, 1.32 | 0.4 |
| Other | 1.27 | 1.23, 1.31 | <0.001 |
| Unknown | 5.28 | 5.11, 5.46 | <0.001 |
| Hispanic | |||
| No | |||
| Other/unknown | 2.57 | 2.48, 2.67 | <0.001 |
| Yes | 1.48 | 1.43, 1.53 | <0.001 |
| Sex | |||
| Female | |||
| Male | 0.88 | 0.86, 0.90 | <0.001 |
| Other | 4.71 | 3.25, 6.66 | <0.001 |
| Age | 1.01 | 1.01, 1.01 | <0.001 |
| CCI | 1.10 | 1.09, 1.12 | <0.001 |
| Number of diagnoses | 1.00 | 1.00, 1.00 | 0.040 |
| Encounters | 1.01 | 1.00, 1.01 | <0.001 |
Discussion
To our knowledge, this is the first study to analyze the epidemiology of medical record duplication. We found a medical record duplication rate of 2.9% and showed that there was a correlation between duplicate records and both health care utilization and missing demographic data on initial registration. As expected, “sicker” patients (higher CCI), females, and those having more contact with the health care system were more likely to have duplicate records. Interestingly, Black race and Hispanic ethnicity were also associated with having a duplicate record, potentially contributing to disparities in health outcomes.
Strengths of the study included a large, diverse sample of over a million patients with a duplicate rate of 2.9%, in line with other institutions. 19 Although difficult to quantify, the academic health care system providing the data grew by acquisition; buying practices in the area across a variety of clinical specialties. Thus, the sample reflects a variety of practice types, clinical specialties, and registration processes.
Weaknesses include a single-institution sample derived from a single EHR (Allscripts); thus, the generalizability of results to other EHR systems and institutions is unknown. Specifically, previous studies found highly variable prevalence of likely duplicate records at different institutions ranging from 0.16 to 15.47%. 19 In that study, neither the specific approaches implemented to control duplicates nor the overall number of approaches implemented at that institution correlated consistently with the prevalence of duplicates at that institution.
Further, our analysis does not allow us to determine the reasons why duplicates were created. However, the absence of diagnoses from most (87.8%) of the duplicate records suggests that the duplicate records resulted from intersystem communication, rather than duplicate registrations for encounters within the system. For example, if a patient was admitted to an affiliated hospital, the inpatient system sent demographic information to Allscripts, but there would be no diagnoses or clinical data.
Automated identification of duplicate records by the eMPI in our study was probably not perfect (worst-case error rate 2.5%). However, errors were more likely to be false negatives than false positives. Therefore, the prevalence of duplicate records in Hispanic or foreign-borne patients was, if anything, underestimated.
Our data do not allow us to determine the validity of demographic variables. For example, we do not know whether the sex variable is most consistent with biological sex or gender identity and how it was ascertained (e.g., patient self-report vs. assumed by the person entering data). Similar limitations apply to data on race and ethnicity. While it is plausible that, Hispanic ethnicity, Black race, as well as missing or unknown demographic data could increase the likelihood of not being able to find a patient in the EHR system, which then results in a duplicate at a later encounter, causal studies would be needed to estimate the strength of this relationship.
The United States lacks a universal patient identifier (UPI). Thus, countries that have implemented UPIs such as Canada, England, Wales, Spain, and Denmark may have different distributions of duplicate records. However, countries with UPIs still face a variety of challenges such as nonresident individuals who seek health care but lack a UPI. 20 One might hypothesize that racial and ethnic minorities are likely to be overrepresented within nonresident populations and individuals without a UPI are more likely to have a duplicate record compared with individuals with a valid UPI. Therefore, racial and ethnic minorities may be at higher risk of having duplicate records even in countries with UPIs.
Previous studies described correlations between patient data and medical record duplication. Just et al found that a common reason for duplication occurred due to nonstandardized methods in collecting middle names, which resulted in an inability to find previously created patient records. 5 Additionally, Khunlertkit et al found that registrars had difficulty with finding patients with uncommon names, which they suggest could affect patient duplication rates. 11 Previous studies also found that record linkage was more difficult in Hispanic and foreign-born adults. 21
Our findings suggest that duplicate records are not randomly distributed in patient populations and can be used to guide duplicate reduction efforts. For example, our finding that most duplicate records lack diagnoses (i.e., are effectively blank, other than demographic data) suggests that the prevalence of duplicates can be significantly reduced by correcting or clarifying communication between systems. Specifically, admissions to affiliated hospitals should not generate a new record in the outpatient system. Alternatively, if new records are generated, they should be designated as “externally generated” and separated from the production system. Fortunately, once identified such records are easy to reconcile since they are blank and can be easily deleted.
In addition to potentially explaining some disparities in health care outcomes, the epidemiology of duplicate patient records can be useful to institutions that wish to minimize duplicates or to mitigate their negative effects. For example, higher-risk subpopulations can be targeted for interventions that may reduce duplicates such as double data entry to avoid typographical errors, alerts to search for existing records, and using photo or biometric identifiers (e.g., vein pattern matching).
In general, failure to accurately identify patients has been associated with a variety of negative consequences including a variety of “wrong patient” errors including invasive procedure performed on the wrong patient, wrong site, or medication errors. 20 In one study, the vast majority of misidentification cases were associated with similar or identical names. 22 However, the clinical consequences of duplicate records specifically, as opposed to identification errors generally, remain understudied. Although it seems likely that duplicate records can compromise patient safety via various mechanisms, 23 causal effect estimates based on data are lacking. Notably, we previously found that duplicate records were associated with a higher probability of missing abnormal laboratory test results; even if the duplicate record contained no clinical data. 9 Further, duplicate records complicate interoperability efforts. Clinical information regarding a patient with a duplicate record may not be accurately transmitted. For example, information regarding a patient being seen in an emergency department while travelling may not be transmitted or may be incomplete if the patient has a duplicate record at their home institution. Therefore, administrative processes such as patient registration and intersystem communication may have clinical effects and compromise patient safety.
Conclusion
Sicker, older, female or nonmale patients who have more contact with the health care system as well as Hispanic and Black patients or those with missing demographic data upon first registration had higher odds of having duplicate records in EHR systems. Although based on a large sample from a complex, heterogeneous health care system, our findings may not generalize to institutions with different identity management practices. Along with efforts to reduce duplicate records, future work should investigate the relationships between institutional identity management practices, prevalence, and distribution of duplicate records as well as the effect of duplicate records on clinical outcomes.
Clinical Relevance Statement
Duplicate patient records are relatively common in operational health systems. Previous studies suggest that duplicate patient records may be a patient safety hazard due to their association with missed abnormal laboratory results and blood transfusion errors. However, the epidemiology of duplicate records has not been studied. We found four categories of factors to be associated with increased odds of having at least one duplicate record: (1) race/ethnicity: specifically, Black race and Hispanic ethnicity, (2) nonmale sex, (3) missing demographic data, and (4) more contact(s) with the health care system such as comorbid illness (higher CCI, higher number of diagnoses, more encounters) and older age. Our findings suggest that duplicate patient records may contribute to health disparities related to race and ethnicity. The epidemiology of duplicate patient records may also help institutions minimize duplicates by targeting high-risk groups for prevention measures, such as using biometric identifiers.
Multiple-Choice Questions
-
The following demographic factors were associated with increased odds of having a duplicate record except
Older age
Hispanic ethnicity
Black race
Male sex
Correct Answer : The correct answer is option d. Multiple demographic factors were associated with increased odds of having a duplicate record, but male sex was associated with a lower odds (0.88) compared to female sex or unknown sex.
-
Duplicate records
Always contain clinical data
Are very infrequent (<1% at all institutions)
Are associated with missing demographic data
Were demonstrated to cause admissions to hospital
Correct Answer : The correct answer is option c. Missing demographic data such as missing Hispanic ethnicity or race was associated with increased odds of having a duplicate record.
Funding Statement
Funding This work was supported by the National Institutes of Health/National Center for Advancing Translational Sciences grant number UL1 TR003167 and the Reynolds and Reynolds Foundation.
Conflict of Interest None declared.
Authors' Contributions
All authors participated in the problem formulation and experimental design. O.S., A.Z., R.J.A., and T.R.J. analyzed the data. All authors participated in drafting and revising the manuscript.
Human Subjects Protection Statement
This study has been approved by the Committee for the Protection of Human Subjects (the UTHSC-H IRB) under protocol HSC-SBMI-13-0549.
Data Availability Statement
The data underlying this article cannot be shared publicly due to the fact that these data are individually identifiable and represent real-world patients.
References
- 1.Makary M A, Daniel M. Medical error-the third leading cause of death in the US. BMJ. 2016;353:i2139. doi: 10.1136/bmj.i2139. [DOI] [PubMed] [Google Scholar]
- 2.Lippi G, Mattiuzzi C, Bovo C, Favaloro E J.Managing the patient identification crisis in healthcare and laboratory medicine Clin Biochem 201750(10-11):562–567. [DOI] [PubMed] [Google Scholar]
- 3.Kohn L T, Corrigan J M, Donaldson M S.To err is human: building a safer health system National Academies Press (USA)2000. Accessed April 25, 2024 at:https://www.ncbi.nlm.nih.gov/books/NBK225187/ [PubMed] [Google Scholar]
- 4.Palabindala V, Pamarthy A, Jonnalagadda N R. Adoption of electronic health records and barriers. J Community Hosp Intern Med Perspect. 2016;6(05):32643. doi: 10.3402/jchimp.v6.32643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Just B H, Marc D, Munns M, Sandefer R.Why patient matching is a challenge: research on Master Patient Index (MPI) data discrepancies in key identifying fields Perspect Health Inf Manag 201613(Spring):1e. [PMC free article] [PubMed] [Google Scholar]
- 6.Cummins D. Errare humanum est ... but it is also human to prevent errors. ISBT Sci Ser. 2007;2(01):249–252. [Google Scholar]
- 7.Hillestad R, Bigelow J H, Chaudhry Bet al. Identity Crisis: An Examination of the Costs and Benefits of a Unique Patient Identifier for the U.S. Health Care System RAND Corporation; 2008. Accessed April 25, 2024 at:https://www.rand.org/pubs/monographs/MG753.html [Google Scholar]
- 8.McClellan M A.Duplicate medical records: a survey of Twin Cities healthcare organizationsAMIA Annu Symp Proc AMIA Symp. 2009;2009421–425. [PMC free article] [PubMed]
- 9.Joffe E, Bearden C F, Byrne M J, Bernstam E V.Duplicate patient records–implication for missed laboratory resultsAMIA Annu Symp Proc AMIA Symp. 2012;20121269–1275. [PMC free article] [PubMed]
- 10.Cohen R, Ning S, Yan M TS, Callum J. Transfusion safety: the nature and outcomes of errors in patient registration. Transfus Med Rev. 2019;33(02):78–83. doi: 10.1016/j.tmrv.2018.11.004. [DOI] [PubMed] [Google Scholar]
- 11.Khunlertkit A, Dorissaint L, Chen A, Paine L, Pronovost P J. Reducing and sustaining duplicate medical record creation by usability testing and system redesign. J Patient Saf. 2021;17(07):e665–e671. doi: 10.1097/PTS.0000000000000434. [DOI] [PubMed] [Google Scholar]
- 12.Khunlertkit A, Paine L. A human factors approach for root cause analysis: a case of duplicate medical record number. Proc Int Symp Hum Factors Ergon Health Care. 2015;4(01):156–161. [Google Scholar]
- 13.Bittle M J, Charache P, Wassilchalk D M. Registration-associated patient misidentification in an academic medical center: causes and corrections. Jt Comm J Qual Patient Saf. 2007;33(01):25–33. doi: 10.1016/s1553-7250(07)33004-3. [DOI] [PubMed] [Google Scholar]
- 14.O'Neill K A, Shinn D, Starr K T, Kelley J. Patient misidentification in a pediatric emergency department: patient safety and legal perspectives. Pediatr Emerg Care. 2004;20(07):487–492. doi: 10.1097/01.pec.0000136895.99271.86. [DOI] [PubMed] [Google Scholar]
- 15.Duvall S L, Fraser A M, Kerber R A, Mineau G P, Thomas A.The impact of a growing minority population on identification of duplicate records in an enterprise data warehouse Stud Health Technol Inform 2010160(Pt 2):1122–1126. [PubMed] [Google Scholar]
- 16.Joffe E, Byrne M J, Reeder P et al. A benchmark comparison of deterministic and probabilistic methods for defining manual review datasets in duplicate records reconciliation. J Am Med Inform Assoc. 2014;21(01):97–104. doi: 10.1136/amiajnl-2013-001744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Charlson M E, Pompei P, Ales K L, MacKenzie C R. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(05):373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 18.Gasparini A.Ellessenne/comorbidityMarch 31, 2024. Accessed April 25, 2024 at:https://github.com/ellessenne/comorbidity
- 19.McCoy A B, Wright A, Kahn M G, Shapiro J S, Bernstam E V, Sittig D F. Matching identifiers in electronic health records: implications for duplicate records and patient safety. BMJ Qual Saf. 2013;22(03):219–224. doi: 10.1136/bmjqs-2012-001419. [DOI] [PubMed] [Google Scholar]
- 20.Riplinger L, Piera-Jiménez J, Dooling J P. Patient identification techniques - approaches, implications, and findings. Yearb Med Inform. 2020;29(01):81–86. doi: 10.1055/s-0040-1701984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lariscy J T. Differential record linkage by Hispanic ethnicity and age in linked mortality studies: implications for the epidemiologic paradox. J Aging Health. 2011;23(08):1263–1284. doi: 10.1177/0898264311421369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tase T H, Quadrado E RS, Tronchin D MR. Evaluation of the risk of misidentification of women in a public maternity hospital. Rev Bras Enferm. 2018;71(01):120–125. doi: 10.1590/0034-7167-2017-0134. [DOI] [PubMed] [Google Scholar]
- 23.Solomon R P, Goldberg-Alberts A, Marella W M ECRI Institute patient safety organization deep dive: patient identification. 2016. https://www.ecri.org/Resources/Whitepapers_and_reports/PSO%20Deep%20Dives/Deep%20Dive_PT_ID_2016_exec%20summary.pdf https://www.ecri.org/Resources/Whitepapers_and_reports/PSO%20Deep%20Dives/Deep%20Dive_PT_ID_2016_exec%20summary.pdf
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data underlying this article cannot be shared publicly due to the fact that these data are individually identifiable and represent real-world patients.