Analysis of recording and coding of smoking history for patients admitted to a regional hospital

Abstract

It is important that accurate information about the occurrence of smoking is available in hospital records both for treating health professionals and for coders for public health records, given that smoking is the leading cause of premature death in the developed world. 1,2 However, some reports have questioned the accuracy of medical record data and coding in a range of conditions and settings, so we undertook a small study in our institution. 3-5 Participants, methods and resultsA retrospective study was undertaken of 100 consecutive patients discharged from Dubbo Base Hospital (DBH) between June and August 2005. We examined the medical record for the most recent admission to determine what smoking history was recorded, where and by whom, the correlation of smoking history across the recording sites and the correlation with the final coding undertaken by medical records staff for the purpose of reporting of public health statistics. As the study was a review of patient records, participants who were smokers were not contacted directly about smoking cessation. Ethics Committee approval was obtained prospectively.The findings of our evaluation of the recording of smoking history revealed 38 non-smokers, 25 current smokers and 21 ex-smokers, and in 16 cases there was no record of smoking history.Of the 84 patients for whom smoking status was recorded, this was documented in several different locations in the medical record, including in a range of pro formas used at DBH (shown in Table 1).For 24 patients, smoking status was recorded in more than one site. In two cases, there was discrepancy of recorded smoking status during the same admission.The patient care plan pro forma, which was completed by nurses, should be attached in each patient's medical record. For 42 records, this was not attached, while in 23 of the remaining 58 (39.7%), it was attached but smoking history was not recorded. Of the 81 cases in which an emergency department (ED) triage note could be found, smoking history was recorded in only 31 (38.3%). The ward progress note has entries by a range of health care professionals, but information about smoking is entered mainly by the doctor as part of the formal admission note. Of the 89 cases with progress notes, smoking status was recorded in 39 (43.8%). If we would combine ED triage and progress notes as doctors' notes, we can say in only 56% of cases did the doctor record smoking history.Although smoking history was also recorded in a range of other sites in the medical records (Table 1), these sites were relatively uncommon and were unlikely to have been noted either by attending health professionals during the admission or by the coding staff subsequently.The coded smoking status for the 100 files was provided by the Health Information Section of DBH. Fiftynine were not coded (indicating either non-smoker or no information on smoking available), while our review of the records indicated that 38 were non-smokers and for 16 there was no record of smoking status in the medical records (a total of 54...

An audit of the prevalence of recorded nicotine dependence treatment in an Australian psychiatric hospital

“…In a study of Nurse Unit Managers' report of provision of nicotine dependence treatment in all general hospitals in New South Wales, Australia, only 37% of respondents reported that all inpatients had their smoking status recorded, while few (17%) provided brief advice to quit or recommended NRT 13 . Audits of medical records in other Australian general hospitals in both urban and regional settings showed substantial errors in recording smoking status, 14,15 minimal provision of NRT and low prevalence of recorded smoking status on the discharge summary 16 …”

It’s Not That Simple: Tobacco Use Identification and Documentation in Acute Care

“…Decentralized approach is where all nurses (or other clinicians) are expected to provide interventions to their own patients [10]. …”

“…A recent Cochrane review showed that adding a tobacco use question to an EMR increases both the expectation that clinicians will intervene with tobacco and as well as clinician actions for treating tobacco use, and when interventions are provided, cessation increases [6] although meta-analyses show the most effective inpatient interventions are those where the intervention is initiated during hospitalization and followed up for at least one month post-discharge [9]. However, integrating tobacco use questions into EMRs does not ensure that cessation interventions will be provided due, at least in part, to the inconsistency of asking and documentation by different healthcare providers [10]. Moreover, if tobacco use data are being collected, challenges include various errors or inconsistencies such as tobacco status occurring in multiple places in a patient’s chart with discrepant information for the same visit, and usability of the data, which in turn, requires standardizing how tobacco use is coded (e.g., check boxes vs. open text fields) and determining a standard location in patients’ charts so data extraction programs can be designed [10].…”

“…However, integrating tobacco use questions into EMRs does not ensure that cessation interventions will be provided due, at least in part, to the inconsistency of asking and documentation by different healthcare providers [10]. Moreover, if tobacco use data are being collected, challenges include various errors or inconsistencies such as tobacco status occurring in multiple places in a patient’s chart with discrepant information for the same visit, and usability of the data, which in turn, requires standardizing how tobacco use is coded (e.g., check boxes vs. open text fields) and determining a standard location in patients’ charts so data extraction programs can be designed [10]. …”

Tobacco Use among Emergency Department Patients

“…If patients <55 years of age represent a relatively larger proportion of the ED population, the tobacco use rates would be expected to be higher than the general population. Even when tobacco use status is electronically recorded as part of standard practice in the ED, the quality of tobacco use data can be poor and data extraction can be difficult due to the inconsistency of asking and recording status by different healthcare providers, lack of standardization in terms of how tobacco use is coded and where it appears in the chart, and various errors such as tobacco status occurring in multiple places in a patient’s chart with discrepant information for the same visit [9]. …”