Effect of questions used by psychiatrists on therapeutic alliance and adherence Christine Howes Laura Thompson, Rose McCabe 10th October 2017 Christine Howes2015
Background 1 Method 2 Results 3 Conclusions 4 Christine Howes2015
Background 1 Method 2 Results 3 Conclusions 4 Christine Howes2015
Background Psychiatry is inconceivable without clinician questions Questions are the mechanism for: achieving clinical objectives: history taking, reviewing symptoms and deducing diagnostic hypotheses managing the formation of therapeutic alliance – known to impact on treatment decisions and patient adherence (Thompson and McCabe, 2012) in the case of schizophrenia, psychotic symptoms may problematise interaction (McCabe et al., 2002) Christine Howes2015
Background Advice to psychiatrists is often generalised: in general try to use open questions rather than leading questions or closed questions (Burton, 2010) But ‘open’ and ‘closed’ questions encompass numerous question types – with different interactional consequences (Heritage, 2010) Aims: (1) What types of questions do psychiatrists ask patients in routine consultations? (2) Do particular question types predict better therapeutic alliance and treatment adherence? Christine Howes2015
Background 1 Method 2 Results 3 Conclusions 4 Christine Howes2015
Data UK MRC study (collected 2006–2008) (McCabe et al., 2013) 31 psychiatrists from 3 centres Patients meeting DSM-IV criteria for schizophrenia or schizoaffective disorder Consent obtained from 138 (of 579; 40%) Consultations audiovisually recorded 134 verbatim transcripts of verbal dialogue annotated using Dexter (Garretson, 2006) Christine Howes2015
Questions coding Christine Howes2015
Questions coding Question type Example from data Yes/no Do you ever feel someone is controlling your mind? Wh- Where was that done? Declarative So you feel anxious a bit anxious ? Tag You’re on 10mg of olanzapine, aren’t you? Lexical tags I’ll write a letter to your GP , okay? Incomplete Your keyworker is ? Alternative Do you feel better having stopped it or worse? Check Yeah? Wh-in-situ He did what? OCRI Pardon? Christine Howes2015
Questions coding Question type Example from data Yes/no Do you ever feel someone is controlling your mind? Wh- Where was that done? Declarative So you feel anxious a bit anxious ? Tag You’re on 10mg of olanzapine, aren’t you? Lexical tags I’ll write a letter to your GP , okay? Incomplete Your keyworker is ? Alternative Do you feel better having stopped it or worse? Check Yeah? Wh-in-situ He did what? OCRI Pardon? Syntactic Christine Howes2015
Questions coding Question type Example from data Yes/no Do you ever feel someone is controlling your mind? Wh- Where was that done? Declarative So you feel anxious a bit anxious ? Tag You’re on 10mg of olanzapine, aren’t you? , okay? Lexical tags I’ll write a letter to your GP Incomplete Your keyworker is ? Alternative Do you feel better having stopped it or worse? Check Yeah? Wh-in-situ He did what? OCRI Pardon? Lexical Christine Howes2015
Questions coding Question type Example from data Yes/no Do you ever feel someone is controlling your mind? Wh- Where was that done? Declarative So you feel anxious a bit anxious ? Tag You’re on 10mg of olanzapine, aren’t you? , okay? Lexical tags I’ll write a letter to your GP Incomplete Your keyworker is ? Alternative Do you feel better having stopped it or worse? Check Yeah? Wh-in-situ He did what? OCRI Pardon? Intonational Christine Howes2015
Questions coding Question type Example from data Yes/no Do you ever feel someone is controlling your mind? Wh- Where was that done? Declarative So you feel anxious a bit anxious ? Tag You’re on 10mg of olanzapine, aren’t you? , okay? Lexical tags I’ll write a letter to your GP Incomplete Your keyworker is ? Alternative Do you feel better having stopped it or worse? Check Yeah? Wh-in-situ He did what? OCRI Pardon? Open Closed Christine Howes2015
Symptoms Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987) Rated from audiovisually recorded interviews P OSITIVE delusions, sensory hallucinations N EGATIVE blunted affect, emotional withdrawal, alogia G ENERAL anxiety higher scores = greater severity PANSS mean (s.d.) P OSITIVE 13.1 (5.9) N EGATIVE 12.5 (5.8) G ENERAL 28.8 (9.6) Christine Howes2015
Therapeutic Alliance Helping Alliance Scale (HAS) (Priebe and Gruyters, 1993) Psychiatrist rated post-consultation Five items rated 1-10 interpersonal variables e.g. mutual understanding, rapport lower scores = poorer therapeutic relationship Christine Howes2015
Measures and outcomes Adherence to treatment Psychiatrist rated 6 months after consultation (1) Adherence to treatment in general (% of occasions that scheduled appointments were kept and non-medication recommendations were followed) (2) adherence to medication (% of medication taken) 1 > 75 % 2 25 − 75 % 3 < 25 % lower score = better adherence Christine Howes2015
Background 1 Method 2 Results 3 Conclusions 4 Christine Howes2015
Questions used Per consultation Per 1000 words Question TypeTotal Mean (s.d.) Rng Mean (s.d.) Rng YN 2362 16.5 (12.2) 57 12.0 (6.0) 30 Wh- 1700 12.7 (10.4) 63 8.5 (4.8) 23 Declarative 1648 11.0 (8.3) 47 9.0 (8.0) 40 Tag 528 3.9 (4.5) 25 2.3 (2.1) 11 Lexical Tags 496 3.7 (5.2) 29 2.0 (2.2) 11 Incomplete 196 1.5 (1.7) 8 1.1 (1.8) 12 Alternative 159 1.2 (1.5) 10 0.8 (1.2) 9 Check 85 0.6 (1.4) 7 0.4 (1.2) 6 Wh-in-situ 47 0.4 (1.0) 10 0.2 (0.6) 5 OCRI 35 0.3 (0.7) 4 0.2 (0.6) 4 Total 7570 51.7 (32.1) 165 35.0 (16.0) 93 Christine Howes2015
Questions used Per consultation Per 1000 words Question TypeTotal Mean (s.d.) Rng Mean (s.d.) Rng YN 2362 16.5 (12.2) 57 12.0 (6.0) 30 Wh- 1700 12.7 (10.4) 63 8.5 (4.8) 23 Declarative 1648 11.0 (8.3) 47 9.0 (8.0) 40 Tag 528 3.9 (4.5) 25 2.3 (2.1) 11 Lexical Tags 496 3.7 (5.2) 29 2.0 (2.2) 11 Incomplete 196 1.5 (1.7) 8 1.1 (1.8) 12 Alternative 159 1.2 (1.5) 10 0.8 (1.2) 9 Check 85 0.6 (1.4) 7 0.4 (1.2) 6 Wh-in-situ 47 0.4 (1.0) 10 0.2 (0.6) 5 OCRI 35 0.3 (0.7) 4 0.2 (0.6) 4 Total 7570 51.7 (32.1) 165 35.0 (16.0) 93 Christine Howes2015
Correlations with outcomes YN Qs Wh Qs Declaratives r p r p r p PANSS Symptoms General 0.130 0.138 0.152 0.082 0.133 0.131 Positive 0.054 0.052 0.182 0.037 0.028 0.747 Negative 0.182 0.036 -0.008 0.927 -0.010 0.911 Therapeutic Alliance HAS Score 0.030 0.732 0.099 0.259 0.282 0.001 Adherence 0.043 0.636 0.033 0.718 -0.204 0.022 0.00 Correlation is significant at the p < 0 . 05 level 0.00 Correlation is significant at the p < 0 . 01 level Christine Howes2015
Generalised Linear Mixed Model Mixed model statistical method Gamma distribution with a log link function Control for within-individual correlations of psychiatrist Random variables = 3 symptom scales (using a scaled-identity covariance matrix) Fixed variables = 3 question types per 1000 words Christine Howes2015
Therapeutic Alliance more declarative questions = better psychiatrist ratings of therapeutic alliance more wh- questions = worse ratings Christine Howes2015
Adherence more declarative questions = patients more likely to adhere to their treatment Christine Howes2015
Declaratives So what is it about declarative questions? Detailed examination of 210 declarative questions from 30 consultations Checklist type (16/8%) 1 Doctor: Sleeping okay? Verbatim repeats (23/11%) 2 Patient: I’ve had some side-effects. Doctor: You’ve had some side-effects? Inferential/abstract (171/81%) 3 ‘So-’ prefaced inferences (90/53%) Doctor: So you feel a bit anxious? Christine Howes2015
So-prefaced declaratives Patient: It’s just that sometimes in the afternoon I get like, you know, I get the feeling that it’s going to happen to me, I will end up in the hospital. Doctor: Okay. Patient: And, er, Doctor: So you feel a bit anxious? Patient: Yeah. Patient: Yeah, I like to chill out in the house, Doctor, you know, I watch telly and then cook something and then washing and tidy the house up, you know? Doctor: Yeah. So you’re quite happy being on your own? Patient: I’m quite happy, Doctor. Yeah, yeah. Christine Howes2015
So-prefaced declaratives devices for suggesting “something implicitly meant by the client” (Bolden, 2009) display understanding, cooperation and engagement (Antaki, 2008) also serve clinical objectives, as summaries of therapeutic interpretation micro-level formulation (Summers et al., 2017) resource to display sensitivity to the emotional implications of delusions, while avoiding collusion or confrontation Christine Howes2015
Summary Psychiatrists used three question types regularly: yes/no, wh- and declarative Declaratives predicted better adherence and perceptions of the therapeutic relationship wh-questions – associated with positive symptoms – predicted poorer perceptions of the therapeutic relationship Declaratives can propose an understanding of patients’ experiences, in particular their emotional salience Christine Howes2015
Background 1 Method 2 Results 3 Conclusions 4 Christine Howes2015
Recommend
More recommend
