A Letter Intervention in Primary Care Patients to Improve Depression and Anxiety Disorders

In 2004, the UK National Health Service mandated the practice of sending patients copies of all correspondence between clinicians involved in their care. Such a policy is consistent with the principle of autonomy and with the right of the patient to access information about themselves (Somerville, 1994). It is also supportive of a collaborative model of service delivery in which patients are viewed as active agents in their own wellbeing, as opposed to passive recipients of treatment.

Letter interventions have boosted attendance rates for cervical screening (McDowell, Newell \& Rosser, 1989a), blood pressure checks (McDowell, Newell \& Rosser, 1989b), and mammography follow-up (Saywell et al., 2003). They have also resulted in improved adherence to antihypertensive medications and lipid lowering agents (Atthobari, Monster \& de Jong-van den Berg, 2004). However, caution has been urged in relation to the use of letters in some patient subpopulations, such as persons who lack the capacity to consent and patients who have psychiatric diagnoses (Jelley, van Zwanenberg \& Walker, 2002).

Despite these cautions, psychiatric patients react positively to the idea of receiving a copy of the letters containing clinical information that are routinely sent to their GP. At least three quarters of patients who had never received such a letter wished to do so (Dale, Tadros, Adams \& Deshpande, 2004; Marzanski, Musunuri \& Coupe, 2005), and at least 80% of patients who had received a letter supported the practice (Asch, Price \& Hawks, 1991; Nandhra, Murray, Hymas \& Hunt, 2004). Most psychiatric patients wished to continue to receive such letters, and a number commented that "seeing their problems understood and described objectively helped them to gain perspective" (Nandhra et al., 2004, p. 41).

Some psychiatric patients and practitioners, even those who favour letter interventions in principle, have expressed reservations. One commonly expressed concern is over the preservation of patient confidentiality, especially with regard to psychiatric reports that containing highly personal information (see, for example, Marzanski et al., 2005; Nandhra et al., 2004). To date, however, there have been no reported breaches of confidentiality in Britain. In fact, most patients are happy for letters to be mailed to them, as are the carers of patients with dementia (Harris \& Boaden, 2003). Some patients and practitioners are also concerned that the contents of the letter may cause distress (Marzanski et al., 2005; Nandhra et al., 2004). However, the reaction reported by most psychiatric patients themselves was that when they read the letter they did not experience any distress at all. A minority of patients did initially feel distress, but they reported that once they had regained composure they found the letter helpful and were glad they had received it (Nandhra et al.).

An issue that has not yet been clarified is whether receipt of a letter improves adherence to treatment recommendations in cases of psychiatric illness. This is a critical issue; a recent review of adherence to antidepressant medication found drop out rates as high as 33% (Pampallona \& Bollini, 2002). The major aim of the current study is to establish whether adherence to treatment is boosted by a letter intervention. If such an effect is found, it should follow that increased adherence leads to improved reduction in the severity of symptoms, and consequently to improved patient outcomes.

Hypotheses: (1) that receiving the assessment report will improve patient adherence to treatment recommendations; (2) that receiving the assessment report will improve patient outcomes.

Method:

Participants: 200 patients over the age of 18 years who are referred to the Primary Mental Health Team for assessment and who receive primary diagnoses of a depressive or anxiety disorder will participate.

Design: A Group (control v. experimental) x Time (baseline x 1 month x 3 month) x Condition (depressive disorder v. anxiety disorder) will measure (a) patient adherence to treatment recommendations and (b) the severity of symptoms. In addition, qualitative analysis of reasons for suboptimal adherence to treatment recommendations will be performed.

Procedure: Consent will be sought from General Practitioners who refer patients to the PMHEIT for psychiatric assessment (see Appendix A). The patients of consenting GPs will form the participant pool. It is current practice that on referral, patients are given a brochure that explains the function of the PMHEIT and the assessment procedure. This brochure contains a section alerting the patient to the fact that they may be asked to participate in research (see Appendix B). After the clinical interview, patients will be given a participant consent form which contains information on the research (see Appendix C). If consent is obtained, patients will complete the Depression and Anxiety Stress Scales (DASS) and the SF12 questionnaire (a measure of disability). Those whose primary diagnosis is any anxiety or depressive disorder will be selected for inclusion in the trial.

Participants will be randomly assigned to control and experimental groups. For patients assigned to the experimental group, the assessment report will be simultaneously sent to the GP and by registered mail to the patient. Reports will be mailed to referring GPs of control group participants, but will not be mailed to the patient themselves. These patients will be mailed a thank-you letter reminding them of the next follow up call (see Appendix D).

Three weeks after the assessment interview, all participants will be mailed a copy of the DASS and a reminder letter. One week later, a researcher attached to the Primary Mental Health Team will contact the participant by phone. She will ask the participant for his/her responses to the DASS items and will ask a short series of questions regarding adherence to the treatment recommendations. Where adherence is partial or absent, the interviewer will ask the reasons (see Appendix E). This procedure will be repeated three months after the initial interview, when the SF12 will also be administered. On completion of the research, all participants and their GPs will be mailed a summary of the research findings.

Inclusion/exclusion criteria: Criteria for inclusion consist of any of the DSM IV anxiety disorders or depressive disorders. Exclusion criteria will include a primary substance abuse disorder (though co-morbid substance abuse will not be an exclusion criterion), significant personality disorder, psychosis, high suicidal risk or risk of harm to others, or inability to speak and read English to a standard that allows independent completion of the questionnaires. Clinician discretion will also be used to exclude participants for whom a letter would be contraindicated. Assessing clinicians will indicate on their record of assessment if inclusion in the study is not contraindicated.

Randomisation procedures: Patients will be allocated to groups based on a random number sequence recorded in sealed envelopes.

Statistical analyses: A Group (control v. experimental) x Time (Initial x 1 month x 3 month) x Condition (depressive disorder v. anxiety disorder) factorial ANOVA will be performed. Post hoc tests may include Tukey and/or Scheffe analyses. Qualitative analysis of responses to adherence questions will be undertaken using a thematic analysis.

References

Asch, R., Price, J., \& Hawks, G. (1991). Psychiatric outpatients' reactions to summary letters of their consultation. British Journal of Medical Psychology, 64, 3 - 9.

Atthobari, J., Monster, T. B., \& de Jong-ven den Berg, L. T. (2004). The effect of hypertension and hypercholesterolemia screening with subsequent intervention letter on the use of blood pressure and lipid lowering drugs. British Journal of Clinical Pharmacology, 57, 328 - 336.

Dale, J., Tadros, G., Adams, S., \& Deshpande, N. (2004). Do patients really want copies of their GP letters? A questionnaire survey of older adults and their carers. Psychiatric Bulletin, 28, 199 - 200.

Harris, C. \& Boaden, R. (2003). Copying letters to patients: Summaries of 12 pilot project sites. England Department of Health. Available:

Jelley, D., van Zwanenberg, T., \& Walker, C. (2002). Copying letters to patients: Concerns of clinicians and patients need to be addressed first. British Medical Journal, 325, 1359.

Marzanski, M., Musunuri, P., \& Coupe, T. (2005). Copying letters to patients: A study of patients' views. Psychiatric Bulletin, 29, 56 - 58.

McDowell, I., Newell, C., \& Rosser, W. (1989a). Computerised reminders to encourage cervical screening in family practice. Journal of Family Practice, 28, 420 - 424.

McDowell, I., Newell, C., \& Rosser, W. (1989b). A randomised trial of computerised reminders for blood pressure screening in primary care. Medical Care, 27, 297 - 305.

Nandhra, H., Murray, G., Hymas, N., \& Hunt, N. (2004). Medical records: Doctors' and patients' experiences of copying letters to patients. Psychiatric Bulletin, 28, 40 - 42.

Saywell, R. M., Champion, V. L., Zollinger, T. W., Maraj, M., Skinner, C. S., Zoppi, K. A., \& Meugge, C. M. (2003). The cost effectiveness of 5 interventions to increase mammography adherence in a managed care population. American Journal of Managed Care, 9, 33 - 44.

Somerville, M.A. (1994). Labels versus contents: Variance between philosophy, psychiatry and law in concepts governing decision-making. McGill Law Journal, 39, 179 - 199.