In March 2018, an article in New Zealand Doctor, ‘My best friend alcohol’, provided anecdotal reports of alcohol abuse by doctors and nurses and put out a call to “remove the cloak of silence” by examining valid data to identify the extent of alcohol misuse by health professionals. In response, the Health Promotion Agency looked at one aspect of alcohol misuse amongst health professionals, driving under the influence. The integrated Data Infrastructure (IDI) was used to bring together occupation information from the 2013 New Zealand Census and court charges from drink driving.
The Integrated Data Infrastructure (IDI) was used for this analysis. The IDI is a collection of linked administrative and survey datasets which consists of a ‘spine’, which is the central individual-level dataset, and ‘nodes’, which are all the other datasets linked back to the spine. The spine includes the maximum number of people in the target population, and is created using tax, births and visa data (Black, 2016).
Using the IDI, information on occupation from the 2013 New Zealand Census (Census) from Statistics New Zealand and court charges from drink driving from the Ministry of Justice were brought together. This allowed a calculation of the rate of people charged for driving under the influence and the rate of people with offences proven in court related to driving under the influence in health professionals and other professional occupations. Nearly all driving under the influence charges linked to the IDI spine. Very few driving under the influence charges have name suppression (charges with name suppression are not shared with the IDI) (Stats NZ, 2017).
‘Driving under the influence’ offences include driving under the influence of alcohol and/or drugs. However, driving under the influence of drugs makes up only a small percentage of charges.
This analysis includes individuals usually resident in New Zealand who completed the Census on 5 March 2013 and were 18 years or over on that day. Records that did not link to the IDI spine were excluded (5.4%).
Only employed individuals were included in this analysis. 1,884,033 people 18 and over reported being employed in the Census, and 72,651 of them were health professionals, using the Australian and New Zealand Standard Classification of Occupations (ANZSCO) 2006 classification (Australian Bureau of Statistics, 2006). This included 12,498 medical practitioners and 40,506 midwifery and nursing practitioners.
Health professionals are classified as having Skill Level 1 by the ANZSCO. Occupations at Skill Level 1 have a level of skill commensurate with a bachelor degree or higher qualification. At least five years of relevant experience may substitute for the formal qualification. In some instances relevant experience and/or on-the-job training may be required in addition to the formal qualification. Other types of professional occupations, including education professionals and Information and Communication Technology (ICT) professionals, also have Skill Level 1. Because a similar skill level is necessary for the professional occupations, comparisons in rates of drink driving have been drawn between health professionals and other professional occupations. Rates for Police were shown, because they face potentially serious professional consequences for a drink-driving conviction and they may also be more likely to know about the process of alcohol testing, eg, the right to request an evidential blood test.
The court charges data from the Ministry of Justice was searched between January 2012 and January 2015 for charges related to driving under the influence. On average, 86% of medical school graduates are retained two years after graduation (Medical Council of New Zealand, n.d.). By restricting charges to within one year before and two years after the Census there can be confidence that most respondents have not changed occupation.
An indicator of having at least one charge for driving under the influence in the three-year period was created for each individual in the Census. The Australian and New Zealand Standard Offence Classification (ANZSOC) (Pink, 2011) was used to define the indicator, following the method used by the Ministry of Justice (Ministry of Justice, n.d.). Offences for driving under the influence include those in the following ANZSOC groups:
0132: Driving causing death (which relate to alcohol or drugs)
0411: Driving under the influence of alcohol or other substance
1431: Exceed the prescribed content of alcohol or other substance limit.
Note that this indicator counts people charged in court (using the date that the offence occurred). It does not include instances where a person received an infringement notice for driving under the influence of alcohol.
Analysis was done in terms of people charged rather than people convicted for several reasons. Most importantly, a conviction is not the only court outcome that indicates a proven offence. For example, a discharge without conviction requires that the offender was found guilty or pleads guilty (Ministry of Justice, n.d.). In addition, there may be systematic differences in conviction rates between professional groups. For example, courts are aware that there may be serious consequences of a conviction in some professional groups and so may treat these groups differently.
As a sensitivity analysis, repetition of the analysis in terms of people with proven offences was undertaken, which includes discharges without convictions and adult diversion in addition to convictions.
Health professionals had the lowest rate of being charged with driving under the influence in the three-year period (0.27%), compared with the other professional groups (Table 1). All professional groups analysed had a much lower rate of driving under the influence than the average rate for all employed adults (1.26%).
Health professionals had a similar rate of driving under the influence to Police (0.22%). No marked differences within the health professions were found. Medical Practitioners, Health Diagnostic and Promotion Professionals, Midwifery and Nursing Professionals and Health Therapy Professionals all had similar rates.
Analyses of people with proven offences was repeated and found that the pattern of results was essentially the same. This was not surprising, as out of all individuals 18 or over in the Census who had at least one charge for driving under the influence, 98% had at least one proven offence.