Alcohol and Alcoholism Advance Access published online on February 29, 2008
Alcohol and Alcoholism, doi:10.1093/alcalc/agm182
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Improving Alcohol and Tobacco History Taking by Junior Medical Officers
1 Drug Health Services, Royal Prince Alfred Hospital, Sydney South West Area Health Service, Australia
2 Discipline of Addiction Medicine, University of Sydney, Australia
3 Discipline of Medicine, University of Sydney, Australia
4 Discipline of Psychological Medicine, University of Sydney, Australia
5 Royal Prince Alfred Hospital, Sydney South West Area Health Service, Australia
* Author to whom correspondence should be addressed at: Drug Health Services, Page Building, Royal Prince Alfred Hospital, Missenden Road, Camperdown NSW 2050 Australia. Tel.: 61 (2) 9515 7331; Fax: 61 (2) 9515 8970; E-mail: eproude{at}med.usyd.edu.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Aims: We aimed to determine the effectiveness of individual feedback and group feedback in improving recording, assessment, and management of risky alcohol use and of tobacco smoking by Junior Medical Officers (JMOs). Method: Medical records of patients admitted by JMOs were examined for recording of alcohol use, alcohol withdrawal, intervention for alcohol, a consultation with the Drug and Alcohol team, tobacco use, and prescription of nicotine replacement therapy (NRT). In year 1, JMOs from hospital 1 received printed individual feedback on their own and their group's performance, while JMOs at hospital 2 attended a presentation of their group feedback. The following year, they reversed roles. Results: A total of 3025 patient records were examined for 130 JMOs. After individual feedback, the percentage of alcohol histories that were quantified rose significantly, from 69% to 82%. More smokers were detected, and NRT prescribing rates rose significantly. Group feedback showed no change. Logistic regression showed that JMOs were significantly more likely to record an alcohol history if located at the smaller hospital and in first year of hospital practice, if the patient was admitted during business hours, was male, and/or was younger than the median age of 70 years; JMOs were significantly more likely to quantify alcohol consumption after individual feedback, but this had no effect on tobacco history recording. Conclusion: Our study suggests that individual feedback on performance with education about desired standards is effective in improving the quality of recording of alcohol histories by Junior Medical Officers.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Alcohol use was responsible for 4% of the global disease burden in 2000, second only to tobacco use and high blood pressure (WHO, 2002
). It is estimated that the annual economic, health, and social costs to England are £20bn, with 22,000 deaths (Prime Minister's Strategy Unit, 2004). Alcohol excess accounts for more than 72,000 hospital admissions annually in Australia (Chikritzhs et al., 2003). Most of the harms of alcohol occur in nondependent drinkers.
Significant reductions in alcohol use and/or related problems have been reported for brief interventions for problem drinkers identified by screening in many randomized trials in health care settings (Wutzke, 2001; Moyer et al., 2002) Early intervention, as brief as 5–10 minutes, can be effective in reducing alcohol consumption and related problems in nondependent drinkers and can help engage dependent drinkers in treatment (Chick et al., 1985; Aalto et al., 2001).
Admission to hospital presents an opportunity for assessment and appropriate intervention. Between 12% and 36% of hospital in-patients have alcohol problems (Figlie et al., 2000; John et al., 1999; Poon et al., 1994). Further, both tobacco smoking and alcohol consumption have adverse effects on the outcomes of surgery and several medical conditions (Moller et al., 2002; Tonneson et al., 1999; Philpot et al., 1994). Yet, since 1980, several studies have reported that alcohol consumption is not routinely recorded in patient records by medical staff and risky consumption often goes undetected (Barrison et al., 1980; Cerise et al., 1998). Even when risky drinking is recorded, it is often not acted on (John et al., 1999; Dent et al., 1995; Hearne et al., 2002; Aarvold and Crofts, 2002).
A survey of inpatients from a major teaching hospital in Sydney revealed that evidence of an alcohol use disorder (hazardous, harmful or dependent drinking) was self-reported by 12%, using the AUDIT questionnaire (Shourie et al., 2007a). However, alcohol histories were recorded in only 50% of their medical records, and 80% of these did not quantify consumption. No interventions for alcohol were recorded, and 41% of records had no smoking history (Shourie et al., 2007a). Methods for improving detection of drinking problems are needed, so that brief interventions can be appropriately implemented.
Traditional didactic continuing medical education and distribution of educational materials has been shown to do little to change clinical practice in hospitals; for example, in relation to test-ordering (Bauchner et al., 2001; Davis et al., 1995). One Cochrane review of 85 studies showed that audits of clinical practice and feedback were modestly effective in improving professional practice in several settings and across a range of clinical fields (Jamtvedt et al., 2003); however, none of the studies reviewed by Bauchner (Bauchner et al., 2001) nor any of the 26 hospital-based studies in the review of continuing medical education strategies (Davis et al., 1995) involved alcohol.
The primary aim of the present study was to determine whether individual feedback on current clinical practice improved the assessment and management of risky alcohol use in inpatients by Junior Medical Officers (JMOs), i.e., doctors in the first two years of clinical practice. The secondary aim was to determine the effect of this intervention on the recording of tobacco smoking and prescription of nicotine replacement therapy (NRT).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Overview
A quasi-experimental trial (Campbell and Stanley, 1966
) was conducted at two hospitals over 2 years. This examined the effects of individual feedback on their recording and management of alcohol use and smoking, and compared this to the effects of a group educational session, with feedback on the group's performance as a whole. It was hypothesized that individual feedback would be more effective than group feedback. However, we realized that individual feedback would be more labor intensive to provide in routine JMO training. Both hospitals are large urban teaching hospitals in Sydney, Australia, with approximately 65 Junior Medical Officers at each. JMOs are known currently as JMO1 and JMO2 (first and second year of hospital practice, respectively). All JMOs at both hospitals and their supervisors were notified that the study was taking place. However, they did not know when it was to commence. Baseline data on recording of alcohol and smoking history and of interventions relating to smoking or alcohol, were collected at each hospital. In the first year, JMOs at Hospital 1 (H1) then received individual feedback and H2 JMOs received group feedback. In the second year, baseline data were again collected from each hospital and JMOs at that hospital received the opposite intervention (see Fig. 1). Some JMOs in the first year of the study may have been present in the second year, and therefore included in both the individual and the group data collection periods; however it is only known for certain that one JMO took part in both the individual intervention and the group feedback sessions. Follow-up audits were used to assess the effect of the types of feedback on alcohol and smoking history and interventions.
|
Data collection
All the medical records of inpatients in eligible wards who had been admitted by JMOs were examined during 12-week record collection periods; at baseline in years 1 and 2 and postintervention in years 1 and 2. Pediatrics, palliative care, day stay, geriatric rehabilitation unit, sleep center, the birth center and delivery ward, and intensive care and high-dependency wards were excluded (ineligible) wards from the study due to underage patients, extreme illness and practical challenges in collecting data. There are no standard questions on alcohol on any admission forms used by doctors (Shourie et al., 2007b
). Any information is usually entered in the general medical record. During the record audit, information was collected on handheld computers concerning age and sex of patient; type of admission (emergency or elective); ward; clinical specialty; date of admission; hours of admission (business hours, out of hours or not recorded); JMO pager number or name (to allow identification for individual feedback); how alcohol use was recorded (verbatim), any record of intervention for alcohol consumption; use of the Alcohol Withdrawal Scale (AWS); prescription of any withdrawal medication; request for a Drug and Alcohol consultation; record of tobacco use; and, in order to monitor compliance with the hospital policy to offer nicotine replacement therapy (NRT) to all patients who smoke, record of prescription of NRT.
The intervention
The individual feedback
JMOs received printed confidential feedback on their own performance in comparison to that of the overall study group in percentages of alcohol histories recorded and quantified, the number and percentage of interventions for persons drinking over recommended limits, the number and percentage of smoking histories and the number and percentage of cases where NRT was prescribed for smokers. In addition, printed educational guidelines were provided on the desirable minimum standards required.
The group feedback JMOs were invited to a group educational session in which they received feedback of their group's performance as a whole on all the above measures, given by a staff specialist in Addiction Medicine. This was conducted as part of the hospital's routine weekly seminars. Again, guidelines were provided on the desirable minimum standards in history-taking.
|
STATISTICAL ANALYSIS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Data were downloaded and imported into SPSS v12 (SPSS for Windows, 2004
). Alcohol consumption history was re-coded as present or absent, and, if present, as quantified or nonquantified. A quantified history was defined as one from which consumption in grams of ethanol per day could be calculated: "nil alcohol" or "grams = 0" was also coded as quantified. Alcohol consumption was defined as risky if it exceeded Australian NHMRC recommended limits for regular drinking (20 g per day for women or 40 g for men) (NHMRC, 2001). Chi-square test was used to assess changes in percentage of adequate records with quantified alcohol history and tobacco history from baseline to follow-up at each hospital and for each type of intervention for the outcomes of interest; recording of alcohol history, quantified alcohol history, and record of tobacco smoking. Logistic regression (enter method) was used to examine the effects of the individual feedback, group feedback, hospital, JMO gender, or JMO seniority (i.e., JMO1 vs. JMO2), time of admission, and age and sex of patient on each dependent variable, as above. P-values of 0.05 or less were considered significant.
|
ETHICS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
The study was approved by the Human Ethics Review Committee of the Central Sydney Area Health Service. Every JMO was informed of the study each year and provided with the opportunity to decline participation, but none did so.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
A total of 3025 patient records, 2038 at Hospital 1 (H1) and 987 at Hospital 2 (H2) were examined for 130 JMOs (65 at H1 and 65 at H2) across the 2 years of the study. Fewer records were examined at H2 due to the large numbers of long-term and geriatric patients there. Males made up 53% of the patient population; median age was 70 years (mean 65, SD 16.9, range 16–101 years). Across the entire data set the prevalence of drinking over recommended limits was 3%, and of the records where any alcohol history was present, 7%. Twenty percent of these patients had an intervention documented and 19% had a consultation with the Drug & Alcohol team.
After Individual feedback, the rate of recording of any alcohol history remained static at approximately 60%; however, the percentage of alcohol histories that were quantified rose from 69% at baseline to 82% (P < 0.001). The percentage of records with insufficient information to calculate risky drinking decreased from 24% to 19% (P < 0.001). There was no significant rise in the percentage of risky drinkers detected or in recorded interventions. More smokers were detected after individual feedback (20% compared to 17% at baseline; P = 0.038) and nicotine replacement therapy (NRT) prescribing rates rose significantly from 2% to 16% of smokers (P = 0.004) (Table 1).
|
After group feedback, there were no significant changes in recording of alcohol and smoking histories. Drinking status was recorded in 57% of records preintervention and 59% postintervention, and the percentage of alcohol histories that were quantified rose from 75% to 77% (P = 0.39). Smoking status was recorded in 64% of records at baseline and 63% at follow-up. NRT prescribing rates did not change.
Logistic regression showed that factors significantly associated with having an alcohol history recorded were the JMOs being in first year and/or at Hospital 2; the patient being male, aged under the median of 70 years, and being admitted during business hours (Table 2). Where an alcohol history was recorded, the significant predictors of it being quantified were the JMOs having received any feedback that year, and the patients being younger than the median. The effect of individual intervention was slightly greater than that of the group intervention, but the difference was not significant and the confidence intervals overlapped. Factors significantly associated with having a tobacco history present were the JMOs working at Hospital 2, and for the patient, being admitted during business hours, and being male. Numbers of NRT prescriptions were too low to allow further meaningful analysis.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Our study found that provision of individual feedback to JMOs of both their own performance and the performance of their peers in recording of alcohol and smoking histories significantly improved alcohol history recording, in that alcohol consumption was more likely to be quantified. There was a smaller improvement of borderline significance in those who received feedback on their group of JMOs as a whole. While this improvement in recording of meaningful alcohol histories is pleasing overall, JMO performance remained far from ideal. We did not see an increase in detection of risky drinkers or in the rate of provision of interventions to identified drinkers.
Only 119 patients (average 7%) of those who had any alcohol history recorded, either pre- or postintervention, could be classed as risky drinkers based on the recorded levels of consumption or other notes (e.g. "3rd admission for alcohol withdrawal this year"). This suggests underdiagnosis of alcohol problems; compared to the 12% self-reported by patients in a recent survey at one of the study hospitals (Shourie et al., 2007a) and that up to 40% of at-risk drinkers may be going undetected. The majority of risky drinkers (75%) were below the median age of the patient cohort of 70 years, and 85% were men (data not shown). Less than a quarter (n = 24) of those risky drinkers identified were recorded as having received some intervention for their alcohol use; 23 had a visit from the Drug and Alcohol team and 4 also received an intervention from a JMO. Interestingly, more (45; 38%) were monitored with the Alcohol Withdrawal Scale. According to their medical records, the majority (two-thirds) of these patients received no advice in relation to their drinking while in hospital. It is acknowledged that the pressure of time has a great influence, not only on asking but also on recording of information in patients’ notes. It is possible that some patients received advice on their drinking but this was not recorded. However, it seems likely that many risky drinkers still go unrecognized. This can lead to under-recognition of alcohol or nicotine withdrawal and complications of alcohol use or smoking with consequent difficulties in patient management, as well as loss of opportunity for intervention in regard to substance use.
Standardized questions on alcohol use and smoking on all admission forms should increase reporting, but even then the results may not be acted on, pointing to a need for training as well as more informative record keeping. It is known that educational strategies often fail to improve clinical practice; however, interactive programs that use peer discussion and practice sessions for developing skills are more effective than didactic approaches (Davis et al., 1995). One Australian study reported on improvements in alcohol history taking by interns after 3 years of increased undergraduate teaching about alcohol and the introduction of an alcohol and drug unit in the hospital (Gaughwin et al., 2000). Education supplemented by written guidelines and pocket-sized laminated cards about alcohol history-taking was shown in another hospital study to be an effective intervention, measured again by medical record audits (Dent et al., 1995). These techniques resulted in an improvement in JMOs’ performance in alcohol use detection amongst medical patients, but quantified alcohol histories were still absent in half of patients’ notes post-training. However, ours is the first study to demonstrate the effectiveness of individualized feedback to increase the rate at which JMOs record an adequate alcohol history.
JMOs at the smaller hospital (H2) were more likely to take alcohol and tobacco histories. Anecdotally, JMOs at H2 have a lower workload than those at H1, supported by the fact that there is the same number of JMOs in each hospital but many more patients at H1. That could have influenced the time available to record admission notes. Patients admitted during business hours were also more likely to have recorded alcohol and tobacco histories. Both findings suggest that time pressures are an important influence on the adequacy of history-taking. Hospital administration needs to routinely review the impact of work pressures on patient care and the capacity to achieve quality care and preventive medicine.
|
LIMITATIONS OF THE STUDY |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Some JMOs had no admissions during certain data collection periods, due to their postings to excluded wards or country areas and so had no personalized results. Because of this, data analysis was unpaired. This may have diminished the ability of the study to show an effect of the intervention. As mentioned above, some JMOs in the first year of the study may have been present in the second year. We do not know if all the JMOs read their individual feedback.
As an improvement in performance was seen after individual feedback, and to a lesser extent also after group feedback, it could be postulated that this was merely a natural improvement with increasing clinical experience. However, the fact that JMO2s performed less well than JMO1s in alcohol history taking argues against this.
Our study demonstrated that individual feedback led to significantly improved rates of recording of quantified alcohol histories. This finding suggests that individual feedback should be studied further and used more widely in JMO training. The drawback of this particular method is that it was labor-intensive, requiring hours of data collection, analysis, and preparation of individual reports, and therefore could not be widely implemented in its present form. However, the increasing implementation of electronic medical records provides the technology for potential routine implementation. Data can then be extracted and fed back to the medical staff on any aspect of clinical performance. It is also clearly important that senior medical staff give feedback to JMOs on alcohol histories as part of routine clinical supervision. This may not currently occur, as senior doctors routinely graduated before education in assessment and early detection of alcohol problems was routine. It remains a challenge to change entrenched clinical behavior, especially in the presence of major time pressures.
|
ACKNOWLEDGEMENTS |
|---|
Thanks to Drs Lewis Chan and Ted Wu, Directors of Clinical Training, for their support, to the many medical students who collected the data, and to the ward staff of both hospitals. Thanks also to Dr Siva Sivarajasingan of the Health Policy Unit, University of Sydney, for statistical advice. This study was funded by the Australian Education and Rehabilitation Foundation (AER Foundation). The funding body had no part in study design, data collection, management, analysis or interpretation of data, in the preparation, review or approval of the manuscript, nor in the decision to submit the paper for publication.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSSTATISTICAL ANALYSISETHICSRESULTSDISCUSSIONLIMITATIONS OF THE STUDYREFERENCES |
|---|
Aalto M., Seppa K., Mattila P., Mustonen H., Ruuth K., Hyvarinen H., et al. Brief intervention for male heavy drinkers in routine general practice: A three-year randomized controlled study. Alcohol & Alcoholism (2001) 36(3):224–230.
Aarvold A., Crofts T. Missed opportunities? Management of patients with alcohol problems in a surgical ward. Health Bulletin (2002) 60(1):55–61.[Medline]
Barrison I. G., Viola L., Murray-Lyon I. M. Do housemen take an adequate drinking history. British Medical Journal (1980) 281:1040.
Bauchner H., Simpson L., Chessare J. Changing physician behaviour. Archives of Disease in Childhood (2001) 84(6):459–462.
Cabinet Office, Prime Minister's Strategy Unit. Alcohol Harm Reduction Strategy for England (2004) London: Cabinet Office.
Campbell D., Stanley J. Experimental and Quasi-experimental Designs for Research (1966) Boston: Houghton Mifflin Company.
Cerise F., Scarini I. C., Thibodaux R., Cannatella M., Stark S., Brantley P. Alcoholism among indigent inpatients: Identification and intervention by internal medicine residents. Southern Medical Journal (1998) 91(1):27–32.[Web of Science][Medline]
Chick J., Lloyd G., Crombie E. Counselling problem drinkers in medical wards: A controlled study. British Medical Journal (1985) 290:965–967.
Chikritzhs T., Catalano P., Stockwell T., Donath S., Ngo H., Young D., et al. Australian Alcohol Indicators, 1990–2001: Patterns of Alcohol Use and Related Harms for Australian States and Territories (2003) Perth and Melbourne: National Drug Research Institute, Curtin University of Technology and Turning Point Alcohol & Drug Centre.
Davis D., Thomson M. A., Oxman A. D., Haynes R. Changing physician performance: A systematic review of the effect of continuing medical education strategies. JAMA (1995) 274(9):700–705.
Dent T., Shepherd R., London M., Alexander G., Duff C. Education and audit can improve the identification of excessive drinkers among medical inpatients. Health Trends (1995) 27(3):92–7.[Medline]
Figlie N., Pillon S. C., Dunn J., Laranjeira R. The frequency of smoking and problem drinking among general hospital inpatients in Brazil- using the AUDIT and Fagerstrom questionnaires. Sao Paulo Medical Journal (2000) 118(5):139–43.[Medline]
Gaughwin M., Dodding J., White J. M., Ryan P. Changes in alcohol history taking and the management of alcohol dependence by interns at The Royal Adelaide Hospital. Medical Education (2000) 34:170–174.[CrossRef][Web of Science][Medline]
Hearne R., Connolly A., Sheehan J. Alcohol abuse: Prevalence and detection in a general hospital. Journal of the Royal Society of Medicine (2002) 95(2):84–87.
Jamtvedt G., Young J. M., Kristoffersen D. T., O’Brien M. A., Oxman A. D. Audit and feedback: Effects on professional practice and health care outcomes. (2006) [update of Cochrane Database Syst Rev. 2003;(3):CD000259; PMID: 12917891]. Cochrane Database of Systematic Reviews (2), CD000259.
John U., Rumpf H. J., Hapke U. Estimating prevalence of alcohol abuse and dependence in one general hospital: An approach to reduce sample selection bias. Alcohol & Alcoholism (1999) 34(5):786–794.
Moller A., Villebro N., Pederesen T., Tonnesen H. Effect of preoperative smoking in intervention on postoperative complications: A randomised clinical trial. Lancet (2002) 359:114–117.[CrossRef][Web of Science][Medline]
Moyer A., Finney J. W., Swearingen C. E., Vergun P. Brief interventions for alcohol problems: A meta-analytic review of controlled investigations in treatment-seeking and non-treatment-seeking populations. Addiction (2002) 97(3):279–292.[CrossRef][Web of Science][Medline]
NHMRC. Australian Alcohol Guidelines: Health Risks and Benefits (2001) Canberra: National Health & Medical Research Council.
Philpot T. K. The ethics of smoke-free zones: An exploration of the implications and effectiveness of a non-smoking policy as a health-promotion strategy in the context of an orthopaedic trauma ward. Journal of Clinical Nursing (1994) 3(5):307–311.[CrossRef][Medline]
Poon A., Owen J., Gijsbers A. J. Identification of at-risk drinkers in an orthopaedic inpatient population. Australian & New Zealand Journal of Surgery (1994) 64(11):775–779.[CrossRef][Medline]
Shourie S., Conigrave K., Proude E. M., Ward J. E., Wutzke S., Haber P. Pre-operative screening for excessive alcohol consumption among patients scheduled for elective surgery. Drug & Alcohol Review (2007) 26(2):119–125.[CrossRef][Web of Science][Medline]
Shourie S., Conigrave K. M., Proude E., Haber P. Detection of and intervention for excessive alcohol and tobacco use amongst adult hospital inpatients. Drug & Alcohol Review (2007) 26(2):127–133.[CrossRef][Web of Science][Medline]
SPSS for Windows. (2004) Chicago: SPSS Inc. Release 12.0 [program].
Tonnesen H., Rosenberg J., Nielsen H. J., Rasmussen V., Hauge C. Effect of preoperative abstinence on poor postoperative outcome in alcohol misusers: Randomised controlled trial. British Medical Journal (1999) 318:1311–6.
WHO. The World Health Report 2002: Reducing Risks, Promoting Healthy Life (2002) Geneva: World Health Organization.
Wutzke S. The long term effectiveness of brief interventions for hazardous alcohol use: A 10 year followup. (2001) PhD thesis, University of Sydney.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||