Abstract
Cigarette smoking can markedly increase the risk of a number of significant medical problems including lung, oral and other cancers, cardiovascular and cerebrovascular conditions, chronic obstructive and other lung diseases, ulcers and maternal and foetal complications [1]. Because of these clear consequential linkages, cigarette smoking has perhaps tended to be viewed more as a medical problem rather than a mental health issue.
It has taken 50 years for compulsive cigarette smoking to be widely accepted as reflecting an addiction to nicotine [2,3,4]. Nicotine dependence is now established as being like other drug dependencies: a progressive, chronic, relapsing disorder with a mean age of onset of cigarette smoking around 13–14 years and a level of dependence that is inversely related to the age of smoking initiation [5]. The seven generic diagnostic criteria for substance dependence in the DSM-IV [1] can be used to make the diagnosis of nicotine dependence. Nicotine dependence is, therefore, included in an encompassing, contemporary concept of substance dependence, stemming from work on alcohol dependence in the 1970s [6].
Despite the huge amount of money spent on Government education programs publicising the serious medical consequences associated with chronic cigarette smoking, the rate of adolescent smoking continues to rise in most Western countries. For instance, in the US the percentage of eighth and 10th grade adolescents rose by almost 50% from 1991 to 1995 [7]. In New Zealand, however, where there has been a decrease in tobacco product consumption overall in recent times, at a rate faster than in virtually any other industrialised country, 15–19-year-old New Zealanders have shown a slight decrease in daily cigarette smoking from 1991 to 1995 of 26% to 24% [8].
It has been estimated [5] that approximately one in three adolescents who smoke a few cigarettes develop nicotine dependence and of these, approximately one in three die of tobacco-related disease.
The rate of cigarette use among patients with mental health problems (55–90%) is much higher than that among the normal population (30%) [1]. Particular disorders for which there is a higher rate of cigarette smoking are depression, attention deficit disorder, anxiety disorders, bulimia nervosa, conduct disorder, schizophrenia and of course other substance dependence [9].
These considerations make the issue of identifying cigarette smoking in adolescent mental health populations of critical importance. However, there is a glaring lack of published data from adolescent mental health samples smoking suggesting that cigarette smoking may be a neglected facet of assessment in these clinical settings.
The primary aim of this study was to investigate the routine recording patterns of patients' smoking by clinical staff of an adolescent mental health service over a 3-year period. Two points of particular interest were: the rate of recording a diagnosis of nicotine dependence by clinicians in relation to the recording of a history suggestive of this diagnosis; and investigating whether the rate of diagnosis would rise following attention being drawn to this issue within the clinical team.
Method
The files of all patients who underwent an initial assessment or reassessment at the Youth Specialty Service (Mental Health; YSS) over a 2-month period (1 April to 31 May) in 1996, 1997 and 1998 were examined.
The YSS is a dedicated adolescent mental health service in Christchurch, New Zealand, established in 1994 and comprising a multidisciplinary team consisting of psychiatrists, clinical psychologists, social workers and psychiatric nurses all of whom undertake intake assessments. The YSS protocol for assessments is that all patients referred to the service must have a comprehensive individual assessment interview alongside an interview with parents or caregivers, the findings of which are reviewed in a weekly team meeting. A typed, standardised report, which includes a DSM-IV [1] diagnostic formulation, is subsequently worked up and forwarded to the referrer within 2 weeks of the assessment. The clinical file contains a front sheet of information including demographic data and a summary of the diagnosis(es) made.
As this was a study primarily aimed at investigating the routine recording patterns of patients' smoking by clinical staff in a standard clinical setting, it is likely that DSM-IV criteria for nicotine dependence were not necessarily applied at intake assessment in a systematic or comprehensive manner, as is the case for all other diagnoses made. There was no specific smoking questionnaire or semi-structured interview including questions related to smoking routinely used by YSS clinicians.
Information for this study was obtained from the comprehensive report and file front sheet as follows:
Demographics (age, gender and ethnicity of the patient): year of referral and whether a first referral; diagnosis (number of diagnoses, principal diagnosis, specific diagnoses including nicotine dependence, alcohol dependence, cannabis dependence, other drug dependence, major depression and conduct disorder); report (total number of lines in the report, total number of lines dedicated to alcohol and drug issues and number of lines specifically concerning smoking); history of nicotine dependence (defined as current daily use of cigarettes) in the report.
A key point of investigation was the rate of nicotine dependence as determined by the investigators from the report (history of daily use of cigarettes documented) compared with the rate at which clinicians confirmed a diagnosis of nicotine dependence by writing the words ‘nicotine dependence’ either in the report or the front sheet.
At the beginning of 1998, 2 months prior to the 1998 sampling, discussions began within the clinical team regarding nicotine dependence. There was debate about whether making the diagnosis of nicotine dependence is an appropriate activity for a mental health service. However, no new protocols were instituted in terms of requirements for a comprehensive assessment involving smoking or focus on nicotine dependence, nor was there any change in the staff mix. Finally, this present study was conceived and then discussed within the team, following the 1998 2-month time period when clinical files were examined.
Data were entered into SSPS for Windows and analysed using Chi-squared analysis for categorical data and analysis of variance for continuous data.
Results
There were a total of 287 files from the 2-month snapshot (April-May) for the 3 years reviewed. In 1996, there were 84 files, in 1997 108 files and in 1998 95 files. First referrals totalled 232 (80.8%). The rate of first referrals in 1996 (91.7%) was higher than that in 1997 (77.8%) and 1998 (74.7%, x2= 9.29, p = 0.01).
The mean age of patients (at the time of interview) was 14.9 years (range = 12–19 years, SD = 1.3). One hundred and sixty-one patients (56.1%) were male and 126 (43.9%) were female. Two hundred and thirty patients (80.1%) were Caucasian, 32 (11.1%) Maori, nine (3.1%) Pacific Islanders and three (1.0%) were other. There were 13 files that did not include ethnicity. There were no differences in these demographic data across the three years surveyed.
Diagnosis
The mean number of diagnoses per patient was 2.3 (range = 0–6, SD = 1.3). The principal diagnosis recorded from either the front sheet or report and there was a broad range of mental disorders diagnosed. The most frequent principal diagnosis was major depression (23.3%), although disruptive behaviour disorders totalling 27.5% were the largest group (attention deficit disorder 11.5%, conduct disorder 10.8% and oppositional defiant disorder 5.2%). Alcohol and drug disorders comprised 14.9% of the total principal diagnoses, two-thirds of which was either alcohol dependence (5.2%) or cannabis dependence (4.5%). Nicotine dependence was not recorded as a principal diagnosis in any patient in this sample. Other alcohol and drug diagnoses included alcohol abuse, cannabis abuse, polysubstance dependence, opioid dependence and inhalant dependence. Other anxiety disorders included panic disorder, posttraumatic stress disorder, generalised anxiety disorder and obsessive–compulsive disorder. Other disorders included anorexia nervosa, mental retardation, pervasive developmental disorder, dissociative disorder, reactive attachment disorder, gender identity disorder and somatoform disorder. There were 24 files in which no diagnosis was made.
Specific diagnoses
Table 1 shows that the rate of diagnosis (not necessarily principal diagnosis) for major depression, conduct disorder and alcohol dependence remained relatively stable over the 3 years and there was a trend for the rate of diagnosis of cannabis dependence to increase over the 3 years. The startling finding was the dramatic increase in the rate of diagnosis of nicotine dependence from 3.6% in 1996 to 23.7% in 1998.
The rate of diagnosis of common disorders by clinicians in a consecutive sample of adolescent patients assessed during a set two month period over three years, at a specialist youth mental health service (n = 287)
The report
Table 2 shows the data related to size of the clinical report measured in number of lines and specifically the number of lines dedicated to firstly alcohol and drug issues overall, and secondly the number of lines dedicated to history of smoking cigarettes. There were seven files found that did not contain a typed assessment report. The size of the report in total as well as those sections dedicated to alcohol and drug issues and smoking remained constant over the 3 years surveyed.
The number of lines in the clinical reports of a consecutive sample of adolescent patients assessed during a set 2-month period over 3 years, at a specialist youth mental health service (n = 280)
Historical evidence of nicotine dependence
There were no significant differences across the 3 years in terms of the rate at which clinicians recorded daily smoking in the assessment reports, which was taken by the investigators to indicate a probable diagnosis of nicotine dependence. In 1996, the rate was 26.2%, in 1997 32.7% and in 1998 37.4% (Chi-squared for linear trend = 2.5, df = 1, p = 0.12). This contrasts with the dramatic increase in the rate that clinicians recorded a diagnosis of nicotine dependence in the 1998 sampling, as seen in Table 1.
Discussion
This study is the first to our knowledge that reviews the rate of nicotine dependence in an adolescent mental health sample utilising routine clinical data. The key findings were that there was a relatively stable historical record of cigarette smoking recorded in the clinical files in the region of 30–40% across the 3 years sampled, but the rate of formal diagnosis of nicotine dependence by clinicians rose from 3.6% in 1996 to 26.3% in 1998. This rise was in spite of stability over this time period of: size of reports and relevant sections (alcohol and drug history, cigarette smoking history); three other key diagnoses, major depression, conduct disorder and alcohol dependence; and demographic data.
Although there was a trend upwards of the rate of diagnosis of cannabis dependence, the only variable that was statistically different over the time period was the rate of first assessment as opposed to reassessment, with higher rates in 1996 compared with 1997 or 1998. However, the rate of diagnosis of nicotine dependence still increased over three-fold from 1997 to 1998, despite similar readmission rates.
Given the evidence that the samples drawn from the 3 years were essentially the same in terms of clinical profile, the best explanation for the dramatic rise in a diagnosis of nicotine dependence, is clinician behaviour. It would seem highly probably that the specific discussion and ensuing debate within the clinical team about cigarette smoking and nicotine dependence brought about the change in clinical behaviour of the clinicians.
Of interest was the parallel rise of a diagnosis of cannabis dependence, albeit not statistically significant, despite any specific focus or debate about the relevance of focusing on cannabis use by patients as was the case for nicotine. It is quite possible that by drawing attention to the diagnosis of nicotine dependence, clinicians naturally extrapolated their clinical interest to cannabis or perhaps found it subsequently easier to talk to their patients about their cannabis use after having first addressed nicotine.
The findings from this study suggest that the prevalence of nicotine dependence in this outpatient adolescent mental health service is in the region of 30–40% (going by historical evidence in clinical assessment reports), although the rate at which a formal diagnosis of nicotine dependence lags this by 5–10% at the current time. More systematic research, not reliant on routine data collected by clinicians, is necessary to gain a better estimate of a prevalence statistic.
Smoking rates in 15–19-year-old New Zealanders are the region of 20% [8]. Allowing for the fact that this was a younger sample (12–19 years olds) and the method of determination of nicotine dependence by historical recording is likely to be conservative, the rate of nicotine dependence in the region of 30–40% is clearly very high. This finding underlies the importance of not neglecting investigation of cigarette smoking and nicotine dependence as has traditionally been the case in adolescent mental health settings. This is particularly so considering the association of early (adolescent) onset with a more severe and progressive course [5] and the predictable risk in decreasing life expectancy associated with cigarette smoking in any event, let alone adolescent onset.
Making the diagnosis is one thing; intervening is another. There is a dearth of information about effective intervention with nicotine dependent adolescents in general, let alone those who additionally have significant mental health problems. It is clear that nicotine replacement therapy is a useful strategy in adults [10] and there are recent developments in effective pharmacotherapy of nicotine-dependent adults with antidepressant medication, particularly where there is accompanying mood disorder [11,12]. Neither of these has been studied in adolescent mental health patients to date.
Given the data on brief interventions in people with alcohol problems [13], brief interventions are also likely to be helpful in mild cases of nicotine dependence and indeed for those adolescent smokers who have not yet become dependent. Brief interventions are also likely to be useful as a first step for patients who have great ambivalence with reducing or ceasing nicotine use. More intensive treatment including pharmacological strategies can be reserved for the next step.