A Controlled Trial of Extended Brief Interventions in Alcohol Dependent Patients

Hypothesis Extended Brief Interventions (EBI) delivered to alcohol-dependent patients in a hospital setting by an Alcohol Specialist Nurse (ASN) will be effective when compared to usual care; in reducing overall alcohol consumption and improving on the standard measures of alcohol dependence.

Overall aim To test the clinical and cost effectiveness of extended brief interventions for individuals with alcohol dependence in an acute hospital setting.

Objectives:

1. Determine whether brief interventions reduce overall alcohol consumption.
2. Determine whether brief interventions are effective in the treatment of alcohol dependence.
3. Determine whether brief interventions reduce the length of stay in hospital
4. Determine how patients feel about being approached in an opportunistic manner for screening for alcohol-related problems and subsequently being asked to be part of a research trial (Refer to Part B of the protocol. Qualitative analysis of patient views and experience).
5. Determine the cost effectiveness of brief interventions for alcohol dependence. (Part C. Cost effectiveness analysis) Part A. Clinical Intervention Methods Design Randomized Controlled Trial Outcome measures Primary • Severity of Alcohol dependence (SADQ)

Secondary

• Alcohol Use Disorders Identification Test (AUDIT)
• Measure of Dependence (Leeds Dependency Questionnaire) [24]
• Quantity and frequency of alcohol consumption in UK units per drinking day.
• Readiness to Change Questionnaire (RTCQ)
• Number of ED attendances 6 months pre and 6 months post treatment/control.
• Hospital length of stay sum days in hospital 6 months and 6 months post treatment/control.
• Number of hospital admissions 6 months pre and 6 months post treatment/control.
• Length of stay for initial treatment in days.
• Biochemical indicators of co-morbid conditions. Gamma-glutamyl transferase (GGT), Alanine aminotransferase (ALT) and mean corpuscular volume (MCV), when available.

Inclusion Criteria

1. Patients with a score equal to or greater than 16 on the AUDIT screening tool.

Exclusion Criteria

1. Patient's known to be intravenous drug abusers.
2. Patients unable to give informed consent.
3. Patients whose medical co-morbidity requiring frequent or long term hospital admission. These will include patients who are/have;
4. Comatose /unconscious
5. End Stage Chronic Liver Disease
6. Dementia
7. Congestive Cardiac Failure stage III or IV
8. Acute Pancreatitis
9. Head Injury with abnormalities on CT scan
10. Cerebral Vascular Accident (CVA)
11. Chronic Obstructive Pulmonary Disease (COPD)
12. Other chronic conditions where an admission of 1 month or more is required.
13. Patients known to be pregnant
14. Patient's presently recruited to any other research trial
15. Patients currently or recently (last 6 months) engaged in structured alcohol treatment Setting

Southport and Ormskirk district general hospital. This is an acute NHS Hospital Trust with an Emergency department (ED) based within the North West region of England and currently has no input available for patients with alcohol-related admissions or attendances.

Sample Size

Based on a previous study by ourselves on dependent drinkers receiving interventions from an ASN, it is expected that 55% of such patients will display a fall in SADQ score between baseline and 12 weeks' follow up. Another confounder in this area of research is the phenomenon of natural recovery (NR). The natural recovery rate over this time period expected in the control group is expected to be no more than 25% (the literature ranges from 12% with a treatment population [20] up to 35% within a general lifetime population [22, 23]). In order to detect this difference between the groups (55% vs. 25%) with 90% power at the 5% significance level, approximately 65 patients are required in each group. In order to allow for an estimated 50% drop out rate (observed in previous studies in similar patients), 130 patients will be recruited per group. It is estimated that at least 5 eligible patients will be recruited per week, and thus this target should be achieved within the 12 month recruitment period.

Statistical Analysis A full statistical analysis plan will be written prior to the conduct of any comparative analysis of the treatment arms. The primary analysis will be intention to treat and will compare the two groups of patients in terms of their primary outcomes.

Test scores will be summarised using descriptive statistics, means with 95% confidence intervals (or medians with inter-quartile ranges if non-normally distributed) at baseline, 12 weeks and 6 months. Change from baseline at 12 weeks and at 6 months post treatment will be presented. The hypothesis of no difference between the two treatment arms at 12 weeks and 6 months (separate analyses) will be tested using analysis of covariance (ANCOVA), controlling for baseline measurements . A p-value of 0.05 (5% level) will be used to declare statistical significance and 95% confidence intervals of the estimated effects will be reported. Time by treatment interaction will be assessed. The assumptions that are made when using ANCOVA will be assessed.

All longitudinal data collected will be used in a secondary analysis, with a resulting increase in power. A formal test of a treatment-covariate interaction will be conducted. Exploratory analysis will be conducted as to the impact on any treatment effect of other factors such as gender.

As much information as possible will be collected about the reasons for missing outcome data and this will be used to inform any imputation approaches employed in the analysis.

Recruitment Process Patients will be referred by all healthcare professionals at Southport NHS Trust; the ASN will also visit the Emergency Department and Medical Assessment Unit daily to identify cases. The nurse taking consent from the patient will discuss with the patients named nurse if they agree that the patient is physically and mentally fit to give consent to the study.

If the patient gives consent, they will undergo a full alcohol history and assessment. Those patients not wishing to take part of the study, or those unwilling to consent to randomisation will be given usual clinical care, and may be asked to consider consent for Part B of the study. (refer to Part B of this protocol) Consecutive patients will be screened for alcohol misuse on attendance at the ED department of Southport Hospital, on identification of an alcohol-related problem the clinician will refer to the ASN for further assessment.

Consent All eligible patients will be given relevant verbal and written information about the research and asked to consent. Due to the opportunist nature and design of the study patients will be asked to consent on initial assessment. This is to ensure that the population recruited are representative of the clinical population attending ED either aware or unaware of the role of alcohol in their presentation. Patients will be able to withdraw from the study at any point. If reasons for opting out are volunteered by the patient we will ask consent to use this information.

Randomization Procedure On consent patients will be randomized to either the intervention or control group. Patients will be randomized using sequentially numbered opaque sealed envelopes prepared according to a computer-generated randomisation allocation sequence. Blocked randomisation using randomly varying block sizes (prepared using Stata version 8.2) will ensure equal numbers of patients are recruited into each group. Randomisation services will be provided by an independent statistician at the Centre for Medical Statistics and Health Evaluation, University of Liverpool, and the trial will be run from Southport General Hospital.

Process Following consent, the research nurse will complete the screening and diagnostic tools; a) Severity of Alcohol Dependence Questionnaire (SADQ), b) Leeds Dependency Questionnaire c) Readiness to Change Questionnaire (RTCQ), and d) EuroQoL EQ-5D. The patient will then be randomized to treatment group by the research nurse.

Group 1. Treatment Group Patients will receive an initial assessment and EBI from the ASN. The Intervention will be delivered utilising the FRAMES approach [11] (Figure 2). The most important element of this model is the exploration of patients' perceptions as to the link between their alcohol consumption and presentation.

A maximum of six treatment sessions, lasting 20 minutes each, will take place within the 12 week treatment period (2 hours in total). Follow-up visits will be scheduled 12 weeks and 6 months following recruitment into the study, and will be performed in the outpatients department, at home or by telephone. The assessor will be blinded to treatment allocation.

Patients differ in whether they are completing a pharmacological detoxification or have opted for an alcohol reduction strategy utilising tools such as a drink diary. For patients undergoing pharmacological detoxification abstinence is essential and non-negotiable. Within the FRAMES model patients have a choice of treatment modes and aims. Patients are offered up-to 6 interventions. Assessment and brief intervention takes place on each occasion, however biochemical assessment takes place only on clinical necessity (for example when a change in clinical condition indicates).

Group 2. Control Group The control group will receive the same assessments as the intervention group but they will not receive the EBI. They will receive normal clinical care, which may or may not include referral to a specialist alcohol service. Follow up visits will be arranged 12 weeks and 6 months following recruitment into the study, either in outpatients, at home or by telephone. The assessor will be blinded to treatment allocation.

Data Management Data will be collected prospectively at time points specified;

1. Baseline
2. 12 week when treatment has finished
3. 6 months follow up

Data from the Case Report Form (CRF) which will carry only an anonymised number, will be entered directly on to an Microsoft access database held on a password protected networked computer at Southport General Hospital. Patient information will be kept on a patient log, which will remain in a locked filing cabinet in a clinical room at Southport General Hospital alongside the anonymised case report files.

Follow Up Strategy

1. A 48% failure to attend for follow up has been included in our power calculation to account for the drop-out rate.
2. An intention to treat analysis will be performed on all data
3. Care will be taken to take address and phone details of the patient and off another contact person.
4. Consent will be sought to contact the patients General Practitioner (GP) as a form of contact. Therefore GP details and phone number will be taken down.
5. Patients will be offered a range of follow-up options either in the hospital, on the telephone, or at home.
6. 2 weeks before follow up is due the patient will be contacted by the nurse or trial co-ordinator to see if they will be attending their appointment and where they would prefer this appointment to take place.
7. The day before their appointment they will be contacted as a reminder that their follow-up appointment is the next day.
8. If the patient is unable to be contacted they will be sent a letter asking them to contact the team in order to arrange follow-up.
9. If the patient is unable to be contacted via telephone or fails to respond to a letter for the 6 month follow-up appointment, the research nurse will visit the patients home to determine a) if the patient still resides at that address, and b) if the patient is still willing to undertake a follow-up appointment.
10. If the patient cannot be contacted by the team at this stage, they will be said to be lost to follow-up.

Data Protection will be in accordance with Southport and Ormskirk policies for data governance.

Ethical Consideration

Each group will receive normal clinical care in their respective groups. No patient group will be disadvantaged as clinical care in these patients varies, and we do not know which the most effective treatment is.

Part B. Patient Satisfaction with the Service

Aims This part of the research is designed to capture the patient's experience. It adopts a qualitative design and will focus on gaining the patient's views on three main features of the study; a) of being screened for alcohol-related problems, b) of being invited to take part in a treatment programme, c) subsequent participation in the RCT.

For all groups, consent for interviews will take place on initial assessment, and all groups will be given a semi structured questionnaire. Group 2 and 3 patients, will also undergo a face to face interview. Group 3 patients, will be interviewed again at the end of the intervention phase.

Methods

Design

Qualitative semi-structured questionnaires and in-depth face to face interviews.

The purpose of this part of the research is to provide an in-depth analysis of the views of the patients approached to take part in this study. The study will therefore use qualitative interviews to provide a deeper, contextualised understanding of the patient's experience of taking part in this study and more particularly of the opportunistic nature by which they will have their drinking assessed and of their participation in the RCT. Data will be collected using two methods. 1) Semi structured questionnaires. 2) In-depth interviews utilising a Topic Guide. The topic guide will be designed to elicit information on the following:

• The experience of the opportunistic nature of the screening.
• The experience of being offered and receiving a brief intervention.
• The experience of being identified as alcohol dependent without being included in the intervention arm of the study.

All patients will be required to sign informed consent prior to interview.

Inclusion Criteria

The three sample groups will be defined as;

• Group 1. Eligible for intervention but not consenting to RCT.
• Group 2. Eligible for intervention consenting to RCT and randomised to control.
• Group 3. Eligible for intervention consenting to RCT and randomised to intervention.

Exclusion Criteria

1. Patient's known to be intravenous drug abusers.
2. Patients unable to give informed consent.
3. Patients whose medical co-morbidity requires frequent or long term hospital admission. These will include patients who are/have
4. Comatose /unconscious
5. End Stage Chronic Liver Disease
6. Dementia
7. Congestive Cardiac Failure stage III or IV
8. Acute Pancreatitis
9. Head Injury with abnormalities on CT scan
10. Cerebral Vascular Accident (CVA)
11. Chronic Obstructive Pulmonary Disease (COPD)
12. Other chronic conditions where an admission of 1 month or more is required.
13. Patients known to be pregnant
14. Patient's presently recruited to any other research trial

Sample Size It is estimated that approximately 10 patients in each group will be sufficient to reach thematic saturation as described by Morse

Recruitment Process (See flow diagram Fig 4) The ASN will ask all patients irrespective of their study involvement whether they are willing to speak to a qualitative researcher.

If the patient is willing the patient will be seen by the researcher following their clinical intervention.

The qualitative researcher will determine the patient group for this part of the study.

Group 1. Eligible for intervention but not consenting to RCT. Patients will be asked to complete the PEQ. Those completing the PEQ will be asked to consent to a face to face interview. At a time and place convenient to the patient the QR will conduct a face to face interview.

Group 2. Eligible for intervention consenting to RCT and randomised to control. The QR will obtain informed written consent to conduct an interview. At a time and place convenient to the patient, and as soon as possible after consent to the study is given, the QR will conduct a face to face interview.

Group 3. Eligible for intervention consenting to RCT and randomised to intervention.

The QR will obtain informed written consent to conduct an interview. At a time and place convenient to the patient, and as soon as possible after treatment is completed, the QR will conduct a face to face interview.

Analysis The identification of themes and patterns in the interview data, using such techniques as constant comparative and deviant case analysis will be the predominant method of organizing, coding and categorizing the data. While the constant comparative method is associated with the principles of grounded theory, comparison represents a central analytic process in inductive qualitative analysis that seeks to derive concepts from data. In analysis the researcher is as interested in 'negative', 'deviant' or 'anomalous' evidence as the identification of patterns of response. In this sort of analysis divergence provides an important way of informing and modifying emergent conceptualisations and explanation.

Where necessary, NVivo, computer assisted qualitative data analysis software that stores and retrieves data will be used. Essentially NVivo software enables the data from the interview transcripts to be directly imported from Word files into NVivo as rich text files and coded according to description and meaning into 'nodes.

Part C. Economic Evaluation

Outcome measures

Primary

• Alcohol Use Disorders Identification Test (AUDIT)
• Severity of Alcohol dependence (SADQ)
• EQ-5D

Secondary

• Number of ED attendances 6 months pre and 6 months post treatment/control.
• Hospital length of stay sum days in hospital 6 months and 6 months post treatment/control.
• Number of hospital admissions 6 months pre and 6 months post treatment/control.
• Length of stay for initial treatment in days.

An economic evaluation of the EBI will be performed. The major part of this will be simple comparison of costs in the two arms, measuring the costs of providing the services against any differences in NHS costs between the two arms which might occur as a result of decreased use of all other services, e.g. hospital re-attendances, accidents, GP visits etc regardless of their relationship to the abuse of alcohol. Data will be collected from patient GP records. The perspective will be limited to that of the NHS. The time frame will be over the six months follow-up time of the study. The costs of the intervention will be as identified in the study. The costs of the use of NHS services (ED or GP attendances cost per bed day as an inpatient) will be taken from standard NHS costs published by the University of Kent. The total costs of NHS care for each patient will be based on these costs and on the amount of service utilisation data collected as outlined above. It may be necessary to adjust the analysis if there are differences in baseline SADQ or in co-morbidities.

A separate analysis of NHS costs specifically for alcohol related services will need to be undertaken: although it might be anticipated that all NHS costs will fall in the intervention arm, this cannot be assumed at this point and a fall in NHS services use may only be seen in alcohol related areas, especially in-patients with severe co-morbidities. Whether a particular NHS intervention is related to alcohol or to other cause will require judgment, to be made by two clinical assessors who are blind to the patient allocation. Conversely, a fall in total costs might be due to the play of chance resulting in a decrease in non-alcohol related morbidities over the relatively short time frame of the study.

If the costs (especially total costs but also alcohol only related costs, provided that non alcohol related costs are equal) in the intervention arm are less than in the non-intervention arm, then the economic case for the intervention is made. If the costs in the intervention arm are greater than in the non-intervention arm, it is necessary to undertake a cost effectiveness study, with the same time frame and perspective.