PEDIATRICS Vol. 111 No. 5 May 2003, pp. 976-980
Resident Documentation Discrepancies in a Neonatal Intensive Care Unit
,
,||
* Robert Wood Johnson Clinical Scholars Program, University of Washington, Seattle, Washington
Department of Pediatrics, University of Washington, Seattle, Washington
Division of Biomedical Informatics, University of Washington, Seattle, Washington
|| Child Health Institute, University of Washington, Seattle, Washington
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Context. Medical errors are common and potentially dangerous. Little is known about the role of documentation errors.
Objective. To determine the proportion of resident physician progress notes that contained discrepancies, and to identify predictors of such discrepancies.
Design/Methods. We conducted a retrospective cross-sectional chart review of resident physician progress notes over 40 random days in a 4-month period in a neonatal intensive care unit. Using predetermined criteria, we compared resident documentation of patient weights, medications, and vascular lines to other sources of information and recorded the numbers of documentation discrepancies.
Results. There were discrepancies in 209 (61.7%) notes with respect to weight, vascular lines, or medications. Discrepancies occurred in the documentation of medications in 103 (27.7%) progress notes, of vascular lines in 119 (33.9%) progress notes, and of weights in 45 (13.3%) progress notes. Notes both omitted information regarding medications (18.2%) and vascular lines (13.9%) and documented inaccurate information regarding medications (18.6%) and vascular lines (30.1%). Patients with more medications or vascular lines, and with longer lengths of stay, were significantly more likely to have higher rates of documentation errors.
Conclusions. Daily progress notes written by resident physicians in the neonatal intensive care unit often contain inaccurate, or omit pertinent, information. Alternative means or methods of documentation are warranted.
Key Words: resident • documentation • errors • NICU
Abbreviations: NICU, neonatal intensive care unit • prn, as needed • IRR, incident rate ratio • CI, confidence interval
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Medical errors are reported to be common and potentially dangerous.1 The Institute of Medicine report on medical errors in 1999 claimed that 44 000 to 98 000 patients die each year because of medical errors.2 Studies in Colorado, Utah, and New York State found that >50% of these errors were potentially preventable.3,4 A recent study reported that 5.7% of inpatient pediatric medication orders contained errors, 1% of which were linked to subsequent potential adverse drug events.5 Notably, the highest rate of adverse drug events was in the neonatal intensive care unit (NICU).5 This confirmed previous studies which found that pediatric inpatients were the must susceptible to medication order error, and overdose was the most common type of error.6
Medication errors are not the only cause of error.2,7 Leape at al3 found that the most common types of other preventable errors were technical (44%), diagnostic (17%), and failure to prevent injury (12%). A study of preventable deaths from 12 hospitals found that preventable deaths often reflected errors in management and diagnosis.8 Bedell et al9 found that along with medication errors and toxic effects, suboptimal response by physicians to clinical signs and symptoms was a cause of preventable death. Decreasing error rates is challenging because errors generally are complex, interlinked, and sometimes identified only by a sentinel event.10,11
One area that remains relatively unexplored is that the role of documentation errors.12–14 Previous reports have discussed the types and importance of documentation errors in the care of patients,15 but descriptive data on such errors are lacking. Given that all medical decisions must, at some level, be based on the information at hand, inconsistent or inaccurate information can potentially contribute to faulty decision-making and adverse outcomes. Discrepancies can occur resulting from both omissions (failure to record information that is accurate) and commissions (recording information that is not accurate). In settings where patient stays are lengthy and complicated, informational errors can also be propagated between providers and in the medical record.15 Because of the extended length of stays, patient acuity, and higher levels of medication errors in the NICU, we felt that such a site was especially prone to informational errors.
To understand better the frequency and predictors of informational errors, we conducted a retrospective chart review of all patients admitted to a tertiary care NICU over a 4-month period. Our specific goals were to determine the proportion of resident physician progress notes that contained discrepancies, and to identify factors that were associated with such discrepancies.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study Site
This cross-sectional study was conducted in an academic tertiary NICU with a socioeconomic and racially diverse population. Resident physicians caring for these patients were training in pediatrics, obstetrics and gynecology, anesthesia, and family medicine. Residents rotate monthly in the unit in their first, second, and third years. Residents were responsible for writing daily progress notes on their patients.
The study was performed after obtaining institutional review board approval at the University of Washington.
Definitions
Resident Progress Notes
The required daily resident progress note is intended to summarize the patient’s changes and care plan each day. Standard of care at this institution mandates that resident progress notes be written daily on all patients. The computer-generated form that this institution uses has sections for patient medications, vascular lines, and daily weights. On days when residents dictated interim or discharge summaries, notes are not required.
Documentation Discrepancies
We defined a documentation discrepancy as an occurrence when information on the progress note did not match a predefined reference standard. For this study, we focused on 3 pieces of information: patient weights, vascular lines, and medications. We defined the reference standard for weight to be the nursing flowsheet, which is where the nurses who obtain daily weights on all infants record them. The reference standard for line access was a combination of the nursing flowsheet, where information about fluids administered to the patients was recorded, and the nursing daily assessment sheet, where the nurses recorded daily what lines were in the patient. The nurses are responsible for documenting such information. We defined a line as an intravenous line, peripheral or central venous catheter, umbilical vein or artery catheter, or arterial line.
For medications, the reference standard was the medication administration record, which was a computer printout from the pharmacy detailing which medications were dispensed, and updated by nurses when medications were given. We excluded vitamins and dietary supplements, as these were not consistently listed in the medication section of notes. We also did not include medications given in hyperalimentation. We defined a patient to be receiving a medication if it had been started before 8 AM on the day a progress note was written. We did not include medications that were started or discontinued later than 8AM each day. We also did not include as-needed (prn) medications that the patient did not receive that day. If, however, a patient was given a prn medication on a given day, such as an albuterol nebulization, then we counted that as a medication the patient was currently prescribed that day. Topical prn medications, such as moisturizers or diaper creams, were not included. Drip medications were counted as medications. In each case, residents were not penalized if it was possible that they were correct at any time of the day.
If a medication was correctly documented by name (patient was receiving the medication and the medication was correctly recorded on the progress note), the dose was checked for accuracy as well. If the dose for a drip medication recorded on the note was a dose being given at any time that day, it was counted as correct. Doses of medications that were omitted or incorrectly added were considered in further analyses.
We counted a line in a patient if that patient had it by 8 AM. If a line was discontinued after a note was written, the progress note was deemed accurate, whether it was recorded or not.
Medication and vascular line documentation discrepancies were divided into 2 groups: (Table 1) errors of omission occurred when a resident did not record on the progress note a medication that appeared on the pharmacy medication administration record or a line that the patient had in place; 2) errors of commission occurred when a resident did record on the progress note a medication that was not on the pharmacy medication administration record or a line that the patient did not have in place. Errors of omission could not occur when the patient was prescribed no medications or had no lines, so these patients were excluded when calculating percentages of notes with this type of discrepancy. All notes were included when calculating errors of commission.
|
Because the daily weights were sometimes recorded after notes had been written, we accepted as correct any weight that exactly matched, to the number of decimal places written, the weight recorded on the day of the note, or the previous day. Again, in each case, the progress note was deemed accurate if the information was correct at any time of the day.
Progress Note Selection
We identified all patients admitted to the NICU during a 4-month (112-day) period from November 14, 2000 to March 14, 2001. We randomly chose 40 days over that period, using Microsoft Excel’s random number function (Microsoft, Redmond, WA). Using patients’ dates of admission and discharge, we identified patients who would have been inpatients on those 40 days and would therefore be expected to have resident progress notes written.
We excluded all admission notes as their format differs significantly from progress notes. We also excluded all dictations that occurred in lieu of progress notes on days of discharge or resident switch days.
Progress Note Abstraction
One physician (A.E.C.) trained a data collector, who was a medical student (E.O.). Each eligible note was reviewed for errors in documentation for weight, lines, and medications. If a note was missing, every effort was made to see if a dictation had been performed on that day, or if the patient was intermediate status. Another author (A.E.C.), blinded to E.O.’s assessment, reviewed 30% of the notes to determine the reliability of collected information. All questions about errors were resolved by consensus of 2 authors (A.E.C. and E.O.).
Analysis
We report documentation discrepancies both by the numbers of notes and by the numbers of discrepancies per note. We also performed regression analyses of discrepancy rates adjusting for a number of covariates using Poisson regression with robust error estimates clustered by patient to control for the nonindependence of notes, because each patient could have had >1 note included in our study.16 This allowed us to calculate incident rate ratios (IRRs) for each covariate. We chose covariates that we thought a priori would increase the likelihood of incurring a documentation error. These covariates included the total number of medications per patient, the number of lines per patient, day of the note (weekend or weekday), the corrected gestational age (estimated gestational age at birth plus day of life), and number of days the patient had been hospitalized. Calculations were performed using the STATA statistical package (STATA Corporation, College Station, TX).
In the 30% of charts reviewed independently by 2 people, the data collector and physician reviewer had excellent agreement for notes with errors in the documentation of weight (99% and 
= .96), lines (95% and = .89), and medications (91% and = .78).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
During the 4-month study period, there were 83 admissions representing 1330 total patient days. The 40 randomly selected days included 339 notes that met inclusion criteria. No progress notes were missing from the chart. These randomly chosen days were distributed over all of days of the week (Sunday, 7; Monday, 6; Tuesday, 4; Wednesday, 6; Thursday, 7; Friday, 6; and Saturday, 4) and over all 4 months of data collection (months 1–7, months 2–11, months 3–11, and months 4–11).
There were discrepancies in 209 (61.7%) notes with respect to weight, lines, or medications (Table 2). Discrepancies occurred in the documentation of medications in 103 (27.7%) progress notes, of lines in 119 (33.9%) progress notes, and of weights in 45 (13.3%) progress notes. In 43 of those weight discrepancies, the weight was recorded inaccurately; in the other 2, the weight was not recorded at all.
|
Medication documentation omissions occurred in 39 notes (18.2%) of the 214 progress notes written for patients on medications at the time of progress note writing (Table 1). Medication documentation commissions were noted in 63 (18.6%) of the 339 notes. In notes where the patient was on at least 1 medication and at least 1 medication was correctly recorded, 192 patients (93.7%) did not have a dose recorded and 8 patients (3.9%) had an incorrect dose listed.
Line documentation omissions occurred in 34 notes of the 245 progress notes written for patients with lines in place at the time of progress note writing (13.9%) (Table 1). Line documentation commissions occurred on 102 (30.1%) of the 339 notes.
A number of factors were found to be associated with the rates of medication documentation discrepancies (Table 3). The rate of medication documentation omissions was found to increase significantly for each additional medication the patient was on (IRR: 1.90; 95% confidence interval [CI]: 1.48–2.43). The rate of medication documentation commissions increased significantly for each day the patient was hospitalized (IRR: 1.03; 95% CI: 1.01–1.06) and for each vascular line the patient had in (IRR: 2.29; 95% CI: 1.47–3.56). Although the point estimate of the IRR for weekend versus weekday notes was substantially <1, this observation did not achieve statistical significance (IRR: 0.56; 95% CI: 0.27–1.12).
|
Factors were also found to be associated with the rates of line documentation discrepancies (Table 3). The rate of line documentation omissions increased significantly for each line the patient had in (IRR: 3.18; 95% CI: 1.74–5.82). The rate of line documentation commissions decreased significantly for each line the patient had (IRR: 0.44; 95% CI: 0.27–0.75).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Discrepancies in resident documentation of weight, medications, or lines occurred in nearly two thirds of resident progress notes in a NICU. Errors of medication documentation omission and commission occurred in approximately the same proportions, but overreporting of lines was more than twice as common as underreporting of lines. Notably, as infants had more lines and medications, the number of discrepancies increased. We also found that length of hospital stay was predictive of documentation inconsistencies. This suggests that the more complicated the patient, the greater the potential for documentation errors, a finding that is not surprising but nevertheless concerning because accurate information may be of greater importance for acutely ill patients. This may also represent propagation errors where repeated transcriptions of previously documented information lead to errors that are subsequently not checked.15 This is particularly true in units using a computer system to assist in progress note generation, like the one studied here. Notes written on patients with more vascular lines had fewer line documentation commissions. As there are a limited number of vascular lines that can be used, this is not surprising. Notes written on weekends trended toward less medication errors. It is not clear why this was the case, and actually runs contrary to our a priori hypothesis. One possible explanation is that cross-covering residents are more conscientious in their documentation because they do not know the patients well.
This study has some limitations that warrant consideration. This was a retrospective review, and therefore could not allow clinical correlation of errors. A prospective design may have been more able to do so. It was conducted in a single teaching institution, and therefore the results must be generalized conservatively. The extent to which this institution is representative of others is not known. Notably, even within this single institution, residents were drawn from several programs covering several specialties. Other institutions not using a computer system to help generate progress notes may be more or less likely to have documentation discrepancies. Some errors of commission may also have been attributable to the database system used by the residents to record information over time. This is offset by a probable reduction in documentation errors of omission and an improved legibility over handwriting. We also chose only a few aspects of the chart to review for documentation discrepancies. There may be other aspects of charting that some view to be more or less important. Our choice of reference standards may also be imperfect. We chose them as standards for further documentation, not as reference standards of actual occurrence. For instance, even if the number the nurse recorded was not always accurate, for the purposes of resident documentation, the recorded number is the weight and should be transcribed as such. No one else routinely obtains weights, and the nursing weights are widely accepted as what should be used for dosing medications. Finally, we have measured a process of care. The extent to which these discrepancies in documentation are actually associated with untoward patient outcomes is not known. However, given the importance of accurate information in making clinical decisions in critically ill neonates, inconsistent or inaccurate information must be interpreted as a potential source of error. Of note, the resident progress note is a secondary document to the attending progress note, mitigating some of the potential for error. Many residents, however, depend on their notes when making decisions, especially when on call at night. It is surely hard to believe that these inconsistencies are desirable. Such discrepancies only cloud the issues of patient care, especially as patients move further and further from the time of documentation.
Discrepancies in documentation are not surprising to any who have worked in the medical setting. The sheer amount of data that must be processed every day is daunting, and every time information is retranscribed it is subject to error.17 The amount of information that is currently expected to be documented is increasing dramatically for legal and medical reasons.18,19 One might expect, based on our findings, that the amount of discordant information in resident progress notes will therefore increase as well.
Our findings have implications for providers, as well as hospital administrators and policymakers. For providers, these findings should raise awareness of the level of documentation discrepancies occurring in daily charting. They should also lead providers to question which parts of the medical record are the most accurate when making medical decisions. For administrators, these findings make a case for system change to prevent discrepancies. For policymakers, these findings should bring into question whether regulations requiring redundant, repetitive documentation of data are in the best interest of patients.
If we require many medical personnel to document their contributions to patient care, even when such contributions overlap, we increase the risk of documentation errors. This is especially true for our most acutely ill patients, who will have the most providers over time. To decrease documentation errors, we should either consolidate the medical record through the elimination of duplicate charting, or implement measures that will reduce the amount of discrepancies in charting.
Electronic information systems may potentially reduce redundant data transcription, and therefore decrease documentation errors.20 Although many systems are in use throughout the country,21 few have been studied to see if they actually lead to a reduction of errors. Because these systems are expensive and often not easy to install, learn, and use, they must be studied before they are used.
|
ACKNOWLEDGMENTS |
|---|
Support for Dr Carroll was provided by the Robert Wood Johnson Foundation.
|
FOOTNOTES |
|---|
Received for publication Jun 10, 2002; Accepted Sep 12, 2002.
Address correspondence to Aaron E. Carroll, MD, Robert Wood Johnson Clinical Scholars Program, H-220 Health Sciences Center, Box 357183, Seattle, WA 98195-7183. E-mail: acarro{at}u.washington.edu
The views expressed within this article are those of the authors and do not necessarily represent the views of the Robert Wood Johnson Foundation or the University of Washington.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
1. Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med.1991; 324 :370 –376[Abstract]
2. Kohn LT, Corrigan J, Donaldson MS. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 2000;xxi: 287
3. Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients. Results of the Harvard Medical Practice Study II. N Engl J Med.1991; 324 :377 –384[Abstract]
4. Thomas EJ, Studdert DM, Burstin HR, et al. Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care.2000; 38 :261 –271[CrossRef][Web of Science][Medline]
5. Kaushal R, Bates DW, Landrigan C, et al. Medication errors and adverse drug events in pediatric inpatients.
JAMA.2001; 285
:2114
–2120
6. Folli HL, Poole RL, Benitz WE, Russo JC. Medication error prevention by clinical pharmacists in two children’s hospitals.
Pediatrics.1987; 79
:718
–722
7. Neale G, Woloshynowych M, Vincent C. Exploring the causes of adverse events in NHS hospital practice.
J R Soc Med.2001; 94
:322
–330
8. Dubois RW, Brook RH. Preventable deaths: who, how often, and why? Ann Intern Med.1988; 109 :582 –589
9. Bedell SE, Deitz DC, Leeman D, Delbanco TL. Incidence and characteristics of preventable iatrogenic cardiac arrests.
JAMA.1991; 265
:2815
–2820
10. Sentinel events: approaches to error reduction and prevention. Jt Comm J Qual Improv.1998; 24 :175 –186[Medline]
11. Lester H, Tritter JQ. Medical error: a discussion of the medical construction of error and suggestions for reforms of medical education to decrease error. Med Educ.2001; 35 :855 –861[CrossRef][Web of Science][Medline]
12. Cradock J, Young AS, Sullivan G. The accuracy of medical record documentation in schizophrenia. J Behav Health Serv Res.2001; 28 :456 –465[CrossRef][Web of Science][Medline]
13. Hassey A, Gerrett D, Wilson A. A survey of validity and utility of electronic patient records in a general practice.
BMJ.2001; 322
:1401
–1405
14. Szeto HC, Coleman RK, Gholami P, Hoffman BB, Goldstein MK. Accuracy of computerized outpatient diagnoses in a Veterans Affairs general medicine clinic. Am J Manag Care.2002; 8 :37 –43[Web of Science][Medline]
15. Christakis DA, Feudtner C. Informational errors. Eff Clin Pract.2000; 3 :301 –304[Medline]
16. Stata Press. Stata Reference Manual, Release 7. Version 7.0 ed. College Station, TX: Stata Press; 2001
17. Andersen SE, Fog D. [Written documentation of drug prescriptions. Accordance between medical records and dispensing records]. Ugeskr Laeger.1998; 160 :4059 –4062[Medline]
18. Moore KJ, Henry LA. Exam documentation just got harder. Fam Pract Manag.1997; 4 :75 –82, 85[Medline]
19. Howell J, Chisholm C, Clark A, Spillane L. Emergency medicine resident documentation: results of the 1999 American Board of Emergency Medicine In-Training Examination Survey. Acad Emerg Med.2000; 7 :1135 –1138[Web of Science][Medline]
20. Kaushal R, Barker KN, Bates DW. How can information technology improve patient safety and reduce medication errors in children’s health care?
Arch Pediatr Adolesc Med.2001; 155
:1002
–1007
21. Menke JA, Broner CN, Campbell DY, McKissick MY, Edwards-Beckett JA. Computerized clinical documentation system in the pediatric intensive care unit. BMC Med Inform Decis Mak.2001; 1 :3[CrossRef][Medline]
PEDIATRICS (ISSN 1098-4275). ©2003 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Prgomet, A. Georgiou, and J. I Westbrook The Impact of Mobile Handheld Technology on Hospital Physicians' Work Practices and Patient Care: A Systematic Review JAMIA, November 1, 2009; 16(6): 792 - 801. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Engel, C. Henderson, J. Fergenbaum, and A. Colantonio Medical Record Review Conduction Model for Improving Interrater Reliability of Abstracting Medical-Related Information Eval Health Prof, September 1, 2009; 32(3): 281 - 298. [Abstract] [PDF]
|
|
|
|
 |
|
 J. R.S. Gordon, T. Wahls, R. C. Carlos, I. I. Pipinos, G. E. Rosenthal, and P. Cram Failure to Recognize Newly Identified Aortic Dilations in a Health Care System With an Advanced Electronic Medical Record Ann Intern Med, July 7, 2009; 151(1): 21 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J E Sheppard, L C E Weidner, S Zakai, S Fountain-Polley, and J Williams Ambiguous abbreviations: an audit of abbreviations in paediatric note keeping Arch. Dis. Child., March 1, 2008; 93(3): 204 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Snijders, R A van Lingen, A Molendijk, and W P F Fetter Incidents and errors in neonatal intensive care: a review of the literature Arch. Dis. Child. Fetal Neonatal Ed., September 1, 2007; 92(5): F391 - F398. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Gray, G. Suresh, R. Ursprung, W. H. Edwards, J. Nickerson, P. H. Shiono, P. Plsek, D. A. Goldmann, and J. Horbar Patient Misidentification in the Neonatal Intensive Care Unit: Quantification of Risk Pediatrics, January 1, 2006; 117(1): e43 - e47. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Garrido, L. Jamieson, Y. Zhou, A. Wiesenthal, and L. Liang Effect of electronic health records in ambulatory care: retrospective, serial, cross sectional study BMJ, March 12, 2005; 330(7491): 581. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S Manzar, A K Nair, M Govind Pai, and S Al-Khusaiby Use of abbreviations in daily progress notes Arch. Dis. Child. Fetal Neonatal Ed., July 1, 2004; 89(4): F374 - F374. [Full Text]
|
|
|
|
|
|
A. E. Carroll, P. Tarczy-Hornoch, E. O'Reilly, and D. A. Christakis The Effect of Point-of-Care Personal Digital Assistant Use on Resident Documentation Discrepancies Pediatrics, March 1, 2004; 113(3): 450 - 454. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. H. Levy Excess Hospitalization Costs Attributable to Medical Injuries AAP Grand Rounds, January 1, 2004; 11(1): 7 - 8. [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. S. Koo Acute Pain Management Journal of Pharmacy Practice, August 1, 2003; 16(4): 231 - 248. [Abstract] [PDF]
|
|
|
|
|
|
G. Suresh Don't Believe Everything You Read in the Patient's Chart Pediatrics, May 1, 2003; 111(5): 1108 - 1109. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
