Teaching junior doctors to manage patients who somatise: is it possible in an afternoon?

Teaching junior doctors to manage patients whosomatise: is it possible in an afternoon?Paul Walters, Andre Tylee, James Fisher & David Goldberg

CONTEXT Many patients in primary care somatisepsychological distress. Training general practitioners(GPs) to manage somatisation has been shown tolead to improvements in their management of thesepatients. However, the training has been intensiveand conducted by psychiatrists, making it impracticalfor widespread use. The aim of this research was todetermine the effectiveness of a teaching package inimproving the ability of GP registrars to managepatients who somatise, when taught by GP vocationalcourse tutors within the constraints of a generalpractice vocational training scheme.

METHODS This was a before-and-after trainingevaluation of GP registrars’ skills. A total of 22 GPregistrars and 6 GP course organisers were recruitedfrom 3 GP vocational training schemes. The GPtrainees had 2 videotaped consultations with trainedactors role-playing patients with somatiseddepression, before and 1 month after training.

RESULTS There was a significant overallimprovement in the ability of GP registrars to managepatients who somatise (mean scores on a 4-pointLikert scale: pre-training 1.4 [standard deviation, SD,0.6]; post-training 2.2 [SD 0.9]; P ¼ 0.002). Generalpractice registrars improved their ability to use anegotiating style of consultation (skill present in 8 ⁄ 22pre-training, 16 ⁄ 22 post-training; P ¼ 0.02) andalso demonstrated more empathy during the�consultation� after training (mean scores on a5-point Likert scale: pre-training 2.3 [SD 1.0];post-training 3.0 [SD 0.8]; P ¼ 0.03).

CONCLUSIONS Using a structured trainingpackage, it is possible for GP vocational course tutorsto successfully teach GP registrars to manage patientswho somatise psychological distress. Given limitedresources for teaching in terms of cost and time, thistraining package could have important implicationsfor training medical staff.

KEYWORDS multicentre study [publication type];evaluation studies [publication type]; humans; male;female; adult; middle-aged; medical staff,hospital ⁄ *education; teaching*methods; education,medical, graduate ⁄ *methods; family practice ⁄*education; somatoform disorders ⁄ *therapy;psychiatry ⁄ *education; physician)patient relations;vocational education; clinical competence; London.

Medical Education 2007: 41: 995–1001doi:10.1111/j.1365-2923.2007.02833.x

INTRODUCTION

Somatisation is the association of medicallyunexplained somatic symptoms with psychologicaldistress and health-seeking behaviour.1,2 Patients whosomatise do not mention psychological symptomsand attribute their symptoms to physical problemswhen consulting their doctors.3 Somatisation iscommon in primary care but general practitioners(GPs) are not confident in managing it and oftenprescribe unnecessary somatic interventions.3–7

A training course has been shown to lead to specificimprovements in the ability of established GPs andGP trainees to manage patients who somatise.8–11

However, the training was delivered by specialists andwas fairly intensive (1 hour per week for 8 weeks).8,12

For such training to be useful and generalisable inprimary care, it needs to be delivered by GPs ratherthan specialists and within the time constraints

postgraduate training

Department of Health Services and Population Research, Institute ofPsychiatry, King’s College London, London, UK

Correspondence: Paul Walters, HSR Department, Box 028, Institute ofPsychiatry, De Crespigny Park, London SE5 8AF, UK.Tel: 00 44 20 7848 5070; Fax: 00 44 20 7848 0333;E-mail: [email protected]

� Blackwell Publishing Ltd 2007. MEDICAL EDUCATION 2007; 41: 995–1001 995

available for teaching in primary care. For GPtrainees, it ideally needs to be delivered by theirclinical tutors or training course organisers within theconstraints of their clinical teaching (i.e. no morethan 1 session of 3)4 hours).

The aim of this research was to determine theeffectiveness of a teaching package on GP registrars’(GP trainees) ability to manage patients who somatisewhen taught by GP vocational training courseorganisers, within the constraints of a vocationaltraining scheme (i.e. 1 · 3–4-hour session) and withno prior training.

METHODS

Study design

This was a before-and-after study design. Anassessment of registrars’ skills in helping patients

reattribute their somatic symptoms was conductedbefore and after being taught with the reattributionskills training pack (see below). The GP registrarsacted as their own controls.

Subjects

General practice vocational training schemes in thesouth-east London area were invited to participate inthe study (n ¼ 10). Three of them agreed to do so(St Georges, Greenwich and Bromley vocationaltraining schemes). All GP registrars from theseschemes were then invited to participate. Accordingto the work of Gask et al.8,13–15 and previous researchby Kaaya et al.,9 20 GP registrars are needed todemonstrate an overall change in the ability of GPregistrars to manage patients who somatise from< 20% to > 75% with 90% power and 95%significance. All 35 GP registrars agreed to participateand pre- and post-assessments were completed for22 (63%; St Georges n ¼ 7, Bromley n ¼ 11,Greenwich n ¼ 4). All GP registrars on the trainingschemes were eligible for inclusion in the study.Six GP vocational course organisers were recruitedfrom the 3 training schemes.

The training package16

The World Psychiatric Association Mental HealthSkills Training Pack was developed in close collabo-ration with GPs especially for use in primary care.It was developed by the Institute of Psychiatry,King’s College London and the National PrimaryCare Research Centre, University of Manchester, incollaboration with the World Psychiatric Association.It consists of 3 components: a booklet explaininghow to run a training session for GPs; a diskettecontaining lecture notes, slides and practice roleplays, and a set of videotapes dealing with a range ofmental health problems in primary care (reattributionskills for somatising patients, chronic fatigue, depres-sion, psychosis in primary care, and dementia inprimary care). As well as providing didactic teachingthrough videotaped interviews with experts demon-strating consultations with patients, each videotapealso teaches and demonstrates a series of practical�microskills� useful in the management of patientswith the above problems. Students then have thechance to practise the microskills through role play.

For this evaluation, course organisers were asked touse the training pack to teach �reattribution skills forunexplained somatic symptoms�. The �reattributionskills for patients who somatise� videotape encour-ages, through the use of role play and discussion, the


Overview

What is already known on this subject

Depression often presents with somaticsymptoms in primary care and generalpractitioners (GPs) feel ill-equipped tomanage it. Intensive training delivered byspecialists can improve GPs’ and GP trainees’skills for managing somatising patients.

What this study adds

It is possible to improve skills for managingpatients who somatise without intensivetraining and within the time constraints of avocational training scheme (i.e. a total of3 hours). The training package can be used byGP tutors with no prior training in its use.

Suggestions for further research

Future research is necessary to determinewhether the training improves clinicaloutcomes in patients. Further research isnecessary to determine whether training givenby GPs to GPs has advantages over that givenby specialists.

996

� Blackwell Publishing Ltd 2007. MEDICAL EDUCATION 2007; 41: 995–1001

development of a set of discrete microskills.12

These skills are: helping the patient feel understood;helping the patient change the agenda of theconsultation from physical complaints topsychological, and helping the patient make the linkbetween his or her physical complaints andpsychological distress.8,12 As the training pack wasdesigned as a stand-alone resource, no training in itsuse was given to the course organisers. Courseorganisers were asked to use the pack within theconstraints of their vocational training course (i.e. for1 afternoon [3 hours]), and to spend only the usualamount of time preparing the teaching.

Pre-training assessment

General practice registrars completed a demographicquestionnaire. They then conducted a 15-minuteconsultation with a trained actor who role-played 1 of2 vignettes: a patient presenting with somatisedstomach pain or a patient with somatised back pain.Registrars were randomly allocated to 1 of the 2role-played scenarios. Actors were trained by apsychiatrist (PW) to ensure the patients theyportrayed all met International Classification ofDiseases-10 (ICD-10) criteria for an underlyingdepressive illness presenting with a somaticsyndrome. The actors were willing to discuss theirdepressive symptoms if asked to. The consultationswere videotaped for assessment purposes.

Post-training assessment

One month after training, the registrars conductedthe second consultation with a different actorrole-playing a patient with the alternative vignette tothat they had been given in the pre-trainingconsultation. These consultations were alsovideotaped for assessment purposes.

Evaluation

The GP registrars rated the usefulness and quality of thetraining package by completing 5-point Likert scales(where 1 ¼ poor and 5 ¼ excellent, and 1 ¼ not at alland 5 ¼ very useful, respectively). The registrars wereasked if they would recommend this pack (Yes ⁄ No)and compared it with their usual teaching sessions interms of complexity, group interaction and the use-fulness of skills taught (on a 5-point scale ranging from1 ¼ much less to 5 ¼ much more).

Course organisers were also asked to rate the qualityof the pack (5-point Likert scale rated from 1 ¼ poorto 5 ¼ excellent), how useful the components of

the pack were (5-point Likert scale rated from 1 ¼ notat all, to 5 ¼ very useful) and whether they would usethe pack again (Yes ⁄ No). They also rated theirconfidence in teaching the management of somatis-ing patients before and after the training (5-pointLikert scale ranging from 1 ¼ not at all, to 5 ¼ very).

Assessment of consultations

The videotaped consultations were rated by a psy-chiatrist (PW) who was blinded to whether they werepre- or post- training. Two rating scales were used.

A scale described by Kaaya et al. was used to rate thespecific skills taught in the package.9 This rating scalehas 3 parts, related to the 3 parts of the reattribu-tion model (i.e. feeling understood, changing theagenda, and making the link). Each skill is measuredand is rated as either present or absent. The skillsmeasured are documented in Table 1. An overallmeasure of registrars’ ability to reattribute somaticsymptoms was obtained by scoring on a 5-point Likertscale (ranging from 1 ¼ poor to 5 ¼ excellent).

General interviewing skills were measured using theConsultation Analysis by Behaviour and Skills(CABS).17 The CABS was developed to measureinterviewing behaviours documented as �desirable� indoctors in their consultations. The measuredbehaviours are all considered as necessary in theconsultation to facilitate the delivery of as muchinformation as possible from the patient and to beperson-centred enough to achieve a sharedunderstanding. The CABS measures doctorbehaviour in terms of desirability set against thecriterion of correct acknowledgement of majordepression in the same interview. It is divided intosections on skills (e.g. doctor actively listens topatient, doctor uses patient’s non-verbal responses),problem definition, key problem identifying andnaming, managing and solving the key problem, andglobal judgements (e.g. empathy, consultation�style�). Skills are measured along 4-point Likert scales(ranging from 1 ¼ rarely used, to 4 ¼ usedfrequently). Empathy was measured on a 5-pointLikert scale (from 1 ¼ the doctor’s response to thefeelings expressed by the patient subtracts noticeablyfrom effective communication, to 5 ¼ the doctor’sresponse allows the patient to express feelings morefully). PW was trained in the use of the CABS by AT.

Analyses

All analyses were conducted using SPSS for WindowsVersion 12.0. Pre-training scores were compared with

997


post-training scores using paired non-parametrictests. Continuous variables were analysed using2-tailed Wilcoxon signed rank tests and categoricaldata with 2-tailed McNemar tests.

RESULTS

Subjects

In total, 22 (63%) GP registrars completed 2 sets ofvideotaped consultations. Ten (45.5%) were maleregistrars. There were no systematic differencesbetween trainees who completed both pre- and post-training assessments and those who completed thepre-training assessment only. The median age was30.5 years (range 26–57 years; mean 33.05, SD7.5 years). Seventeen (77%) had received no mentalhealth training in the previous 2 years and 5 (23%)had attended up to 2 mental health training events inthe previous 2 years. One trainee had previoustraining in somatisation. Nine (41%) had completeda 6-month clinical rotation in psychiatry. None of the

trainees had a special interest in mental health. Themean time between teaching and the second traineeassessment was 4.2 (SD 2.5) weeks.

Six course organisers delivered the training acrossthe 3 vocational training schemes. They had a meanof 14.2 (SD 3.5) years of experience working as GPsand devoted a mean of 12.5% (SD 4.1%) of theirtotal teaching on the training schemes to mentalhealth.

Assessment of the training pack by trainees

Thirteen (59%) registrars thought the quality of thepack was either good or very good. Fourteen (64%)thought the video was useful or very useful and 13(59%) thought the role plays contained in the packwere useful or very useful. Twenty (91%) registrarswould recommend the pack. Eleven (50%) thoughtthe teaching session was more or much morecomplex than usual and 8 (36%) thought it was of asimilar level of complexity as usual. Thirteen (59%)registrars thought there was more or much more


Table 1 Comparison of pre- and post-training skills (present or absent)

Skills present

P-value*

Pre-training

(n = 22)

Post-training

(n = 22)

Feeling understood1 Asks about radiation of pain 3 6 0.452 Asks about typical pain day 1 8 0.023 Asks about associated symptoms 16 21 0.134 Explores health beliefs 18 16 0.695 Explores family and social issues 18 19 1.00Total score (maximum score ¼ 5) 2.6 (1.1)� 3.2 (1.1)� 0.07�

Changing the agenda1 Acknowledges the reality of pain 11 12 1.002 Feeds back result of examination

and tests to patient17 20 0.45

3 Summarises mood symptoms 8 13 0.234 Summarises life events 8 8 1.00Total score (maximum score ¼ 4) 2.0 (1.3)� 2.5 (1.2)� 0.21�

Making the link1 Explains how depression and ⁄ or

anxiety cause symptoms8 17 0.02

2 Practical demonstration ofhow depression causes symptoms

0 1 N ⁄ A

3 Relates pain to the �here and now� 0 1 N ⁄ A4 Relates pain to life events 6 14 0.015 Relates pain to family member 0 1 N ⁄ ATotal score (maximum score ¼ 5) 0.6 (0.7)� 1.6 (0.9)� < 0.01�

Overall negotiating style 8 16 0.02Overall ability to help patientreattribute somatic symptoms

1.4 (0.6)§ 2.2 (0.9)§ < 0.01�

* McNemar’s test (2-tailed)� Wilcoxon signed rank test (2-tailed)� Mean score (standard deviation [SD])§ Mean score (SD) on 4-point Likert scale

998


group interaction compared with their usualteaching sessions, and 13 (59%) thought the skillstaught were more or much more useful than those intheir usual teaching sessions.

Assessment of the training pack by GP tutors

The majority of tutors thought the training pack wasgood quality (mode 4, range 2–5). All thought theteaching pack was at least of some use, and 5 of the 6tutors said they would use the pack again. The GPcourse organisers were more confident in teachingabout somatisation after using the pack, although thiswas not statistically significant (mean scores on a5-point Likert scale: before using the teaching pack2.5 [SD 1.4]; after using the teaching pack 3.8 [SD0.4]; P ¼ 0.06).

Evaluation of before-and-after comparison ofreattribution skills

The changes in reattribution skills following trainingare shown in Table 1.

There was a statistically significant overall improve-ment in GP trainees’ abilities to help patientsreattribute somatic symptoms (P < 0.01). There wasan improvement in registrars’ overall ability to helppatients feel understood, change the agenda, andmake the link between psychological and physicalsymptoms, although there was a statistically signifi-cant improvement only in the latter (P < 0.01). Therewas a statistically significant improvement in thefrequency of GP registrars asking about a typical painday (P ¼ 0.02) and a statistically significant improve-ment in the registrars’ ability to help patientsreattribute somatic symptoms (P < 0.01). Registrarsalso used a more negotiating style when conductingtheir consultation after the training (P ¼ 0.02). Inparticular, trainees explained how depression andanxiety can cause physical symptoms more frequentlyafter training, and also related the patient’s pain tolife events more often after training. Other methodsof helping make the link (practical demonstration,relating pain to the here and now, and relating painto that of a family member) were used too infre-quently for any meaningful inference to be made.

General interviewing skills

After training, the registrars delineated thepresenting complaint in more detail (mean scores ona 4-point Likert scale: pre-training 2.6 [SD 0.8];post-training 3.2 [SD 0.8]; P ¼ 0.01), explained thereason for diagnosing depression (mean scores on a

4-point Likert scale: pre-training 2.6 [SD 1.0];post-training 3.4 [SD 0.7]; P ¼ 0.01), and checkedthe patient understood the meaning of the term�depression� (mean scores on a 4-point Likert scale:pre-training 2.1 [SD 1.0]; post-training 3.1 [SD 1.0];P ¼ 0.01). The registrars also used a more empathicinterviewing style after training (mean scores on a5-point Likert scale: pre-training 2.3 [SD 1.0];post-training 3.0 [SD 0.8]; P ¼ 0.03).

DISCUSSION

We have demonstrated that reattribution skills can besuccessfully taught using a structured teaching pack-age and within the constraints of a general practicevocational training scheme. We have also shown thatGP course organisers can do this successfully, with noprevious specific training on teaching reattributionskills or the management of somatisation. This mayhave implications for other training given by spe-cialists to GP trainees, raising the possibility thattraining given by GPs to GPs may have advantagesover that delivered by specialists. Further research isnecessary to elucidate this. Given limited resources interms of time and money, the use of teachingprogrammes such as this could prove important inhelping trainees acquire the necessary mental healthskills needed for a career in general practice. Thismay be especially important in developing countries,where access to mental health specialists is limitedand primary care professionals have few educationalresources to draw upon.

This study has a number of limitations that need tobe taken into consideration. The sample of GPregistrars was small, but was likely to be representativeof GP registrars in general as no selection criteriaother than being on an approved vocational trainingscheme were used. All GP registrars on the 3 schemesagreed to participate in the research. Although thedrop-out rate was 37%, there was no identifiabledifference between those who completed the pre-and post-training evaluations, and those who com-pleted only the pre-training evaluation based on thepre-assessment information. There are also limita-tions in basing the results on 1 pre- and post-trainingrole-played interview; however, as all trainees weregiven the same standardised role-play scenarios, theresults are a fair comparison of pre- and post-trainingskills. It is also unlikely that the results depend on atime length bias as, in our experience, the skillstaught in the teaching package are quite specific andnot normally learnt over time. In this regard, anychanges in the skills being measured are therefore

999


likely to represent a true effect of the trainingpackage.

It is not possible to say whether these skills generaliseto the clinical environment outside the artificialconstraints of the study, or whether this training willmake any difference to clinical outcomes. Researchhas suggested that isolated educational activities inprimary care may not necessarily lead to improvedclinical outcomes.18,19 However, Morriss and Gaskhave shown that reattribution training amongst GPsleads to fewer beliefs amongst patients that theirsymptoms have a physical cause in the short term.11

Morriss et al. have also shown that reattributiontraining significantly improved function, reducedpsychopathology at 3 months, and reduced theprevalence of depression in those with fixed beliefsabout the origin of their symptoms at 1 month.10

At least 4 studies have demonstrated improvementsin GPs’ reattribution skills following training.8–11

However, the training was fairly intensive and wasdelivered by specialists, limiting its use in primarycare. This study has demonstrated, by comparison,that a brief teaching session, structured around skillslearning and role play, can effect significant changesin skills. This may in part be a result of therelationship between course organisers and regis-trars allowing taught skills to be practised anddiscussed within the context of the clinical encoun-ter, and the clinical relevance of the subject beingtaught.20 As Kaaya et al. found with their moreintensive training programme, this research alsodemonstrated that, although not specifically taught,there were improvements in general consultationskills.9 Trainees were more likely to employ a�negotiating style� post-training, and were also moreempathic during the consultation. It was notpossible in this study to determine the effectivecomponents of the teaching pack, especially interms of improving general interviewing skills.We would, however, speculate that the practicalteaching of specific skills, rather than a didacticpresentation of information is vital, and thatteaching microskills, together with an improvedunderstanding of the subject may lead to animprovement in general interviewing skills.

Gask et al. have also shown that it is possible to trainGP tutors to teach psychiatric interviewing skills.15

They trained GP tutors over 12 weekly sessions, witheach session lasting 1.5 hours. Our results suggestthat, given a fairly structured skills-based teachingprogramme, training GP tutors to teach mentalhealth skills may not be necessary.

Contributors: all authors were involved in the design ofthe study. Data were collected by PW, AT and JF. All authorswere involved in analysing the results and drafting themanuscript.Acknowledgements: we would like to acknowledge thekind assistance of the general practice course organisersand general practice registrars at the Greenwich,St Georges and Bromley general practice vocationaltraining schemes.Funding: this study was funded by the PsychiatricResearch Trust (PRT). The researchers are independent ofthe PRT.Conflicts of interest: AT and DG were involved in thedevelopment of the training package used in this research.Ethical approval: not required.

REFERENCES

1 Reid S, Wessely S. Somatisation and depression. In:Dawson A, Tylee A, eds. Depression: Social and EconomicTimebomb. London: World Health Organization2001;55–61.

2 Gureje O, Simon GE, Ustun TB, Goldberg DP.Somatisation in cross-cultural perspective: a WorldHealth Organization study in primary care. Am JPsychiatry 1997;154:989–95.

3 Bridges K, Goldberg D. Somatic Presentation of DepressiveIllness in Primary Care. In: Freeling P, Downey LJ, MalkinJC, eds. The Presentation of Depression: Current Approaches.London: Royal College of General Practitioners1987;9–11.

4 Weich S, Lewis G, Donmall R, Mann A. Somaticpresentation of psychiatric morbidity in generalpractice. Br J Gen Pract 1995;45 (392):143–7.

5 Kerwick S, Jones R, Mann A, Goldberg D. Mentalhealth care training priorities in general practice.Br J Gen Pract 1997;47:225–7.

6 Ring A, Dowrick C, Humphris G, Salmon P. Dopatients with unexplained physical symptomspressurise general practitioners for somatic treatment?A qualitative study. BMJ 2004;328 (7447):1057.

7 Salmon P, Dowrick CF, Ring A, Humphris GM. Voicedbut unheard agendas: qualitative analysis of thepsychosocial cues that patients with unexplainedsymptoms present to general practitioners. Br J GenPract 2004;54 (500):171–6.

8 Gask L, Goldberg D, Porter R, Creed F. The treatmentof somatisation: evaluation of a teaching package withgeneral practice trainees. J Psychosom Res 1989;33:697–703.

9 Kaaya S, Goldberg D, Gask L. Management of somaticpresentations of psychiatric illness in general medicalsettings: evaluation of a new training course for generalpractitioners. Med Educ 1992;26:138–44.

10 Morriss RK, Gask L, Ronalds C, Downes-Grainger E,Thompson H, Goldberg D. Clinical and patient satis-faction outcomes of a new treatment for somatised

postgraduate training1000

mental disorder taught to general practitioners.Br J Gen Pract 1999;49 (441):263–7.

11 Morriss RK, Gask L. Treatment of patients withsomatised mental disorder: effects of reattributiontraining on outcomes under the direct control of thefamily doctor. Psychosomatics 2002;43:394–9.

12 Goldberg D, Gask L, O’Dowd T. The treatment ofsomatisation: teaching techniques of reattribution.J Psychosom Res 1989;33:689–95.

13 Gask L, McGrath G, Goldberg D, Millar T. Improvingthe psychiatric skills of established generalpractitioners: evaluation of group teaching. Med Educ1987;21:362–8.

14 Gask L, Goldberg D, Lesser AL, Millar T. Improvingthe psychiatric skills of the general practice trainee: anevaluation of a group training course. Med Educ1988;22:132–8.

15 Gask L, Goldberg D, Boardman J et al. Traininggeneral practitioners to teach psychiatric interviewingskills: an evaluation of group training. Med Educ1991;25:444–51.

16 Goldberg D, Gask L, Sartorius N. Training Physiciansin Mental Health Skills. Geneva: World Psychiatric

Association 2001. [Copies of the training package areavailable from [email protected].]

17 Tylee A. Why do general practitioners recognisedepression in only half of their female attenders whoare suffering from major depressive disorder?MD Thesis, University of London 1993.

18 Thompson C, Kinmonth AL, Stevens L et al. Effects ofclinical-practice guideline and practice-basededucation on detection and outcome of depression inprimary care: Hampshire Depression Projectrandomised controlled trial. Lancet 2000;355:185–91.

19 Gilbody S, Whitty P, Grimshaw J, Thomas R.Educational and organisational interventions toimprove the management of depression in primarycare: a systematic review. JAMA 2003;289:3145–51.

20 Hodges B, Inch C, Silver I. Improving the psychiatricknowledge, skills and attitudes of primary carephysicians, 1950–2000: a review. Am J Psychiatr 2001;158(10):1579–86.

Received 24 April 2006; editorial comments to authors18 August 2006, 2 March 2007; accepted for publication10 April 2007


1001

Teaching junior doctors to manage patients who somatise: is it possible in an afternoon?

Documents

Transcript of Teaching junior doctors to manage patients who somatise: is it possible in an afternoon?