- Email: [email protected]
Cognitive therapy for patients with schizophrenia – Authors' reply
Correspondence
Anthony Morrison and colleagues1 provide an interesting perspective on the use of cognitive behavioural therapy (CBT) in patients with schizophrenia spectrum disorders through their pilot study. Although compliance to CBT was reportedly measured, the extent of the adherence to the theoretical construct in actual delivery was not reported in the paper. 1 Although competence of staff providing CBT was deemed adequate, whether and to what degree the therapists used CBT principles during the sessions was not clear. Since this pilot study focused on feasibility, the measurement of such fidelity to psychotherapeutic construct would be important to ascribe the benefits of CBT to the therapeutic process. CBT for patients with schizophrenia spectrum disorders is quite difficult and requires modification of the usual techniques. 2 Engagement is often considered difficult in such patients, so a retention rate of more than 80% by sixth session is quite encouraging. The mean number of sessions (13·3 sessions) suggests that most participants did not complete the planned 26 sessions. Other medication effectiveness studies like CATIE and CUtLASS3,4 also reported similar substantial dropout rates. The authors believe that about 20% of the participants subsequently started antipsychotic drugs. What is not clear from the paper is the proportion of participants in the cognitive therapy group who were prescribed antipsychotics: 10 of 37 would be about 27%. Some patients who had refused medications were subsequently given medications, suggesting that refusal is not definitive. Alternatively, cognitive behavioural measures might make a person think more positively about medications. This study 1 provides valuable insights about a potential treatment where therapeutic alliance might be difficult and drug side-effects might www.thelancet.com Vol 384 August 2, 2014
impair such an alliance. This approach paves the way for more research into other acceptable yet potentially effective alternative treatment modalities for the treatment of schizophrenia in patients who refuse medications.5 We declare no competing interests.
*Sagar Chandra Bera, Siddharth Sarkar [email protected] Department of Psychiatry, AIIMS, New Delhi, India (SCB); and Department of Psychiatry, JIPMER, Puducherry, India (SS) 1
2
3
4
5
Morrison AP, Turkington D, Pyle M, et al. Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial. Lancet 2014; 383: 1395–403. Tai S, Turkington D. The evolution of cognitive behavior therapy for schizophrenia: current practice and recent developments. Schizophr Bull 2009; 35: 865–73. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med 2005; 353: 1209–23. Jones PB, Barnes TRE, Davies L, et al. Randomized controlled trial of the effect on Quality of Life of second- vs first-generation antipsychotic drugs in schizophrenia: Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS 1). Arch Gen Psychiatry 2006; 63: 1079–87. Lee MS, Shin B-C, Ronan P, Ernst E. Acupuncture for schizophrenia: a systematic review and meta-analysis. Int J Clin Pract 2009; 63: 1622–33.
Authors’ reply Our trial1 was not designed to change clinical practice. It was a preliminary trial, which needs to be followed up by a larger, pragmatic multicentre study. It is important not to overinterpret our data, and we explicitly advised against discontinuation of medication. We claimed the trial showed that cognitive therapy was safe and acceptable, not safe and effective. No participants were included on the basis of attenuated
psychotic symptoms (our entry criteria included positive symptoms above a threshold on the positive and negative syndrome scale [PANSS]). Our combined sample had a mean baseline PANSS score of 72, with the recommended cutoffs for mildly ill being 58 and moderately ill being 75; therefore, our sample consists of people with both levels of difficulty. Duration of psychosis and other sample characteristics are reported elsewhere.2 We can clarify that the proportion of patients starting medication in each group was 27%. An investigation of the components of therapy is being undertaken and will be published. Standardised effect sizes (and 95% CIs) for the PANSS subscales are as follows: PANSS positive (–0·45, –0·81 to –0·09), PANSS negative (–0·22, –0·51 to 0·07), and PANSS general (–0·47, –0·78 to –0·16). The claim that missing data is over 50% at follow-up points is based on use of the randomised number as denominator, rather than the number who could have potentially provided outcome assessments given the planned variable length of follow-up (adopted to provide value for money to the funder, allowing maximal recruitment, and the most complete data we could gather in the lifetime of the trial). The proportion of non-missing PANSS data at 9 months (the final point at which an outcome assessment could have been provided by all randomised participants) is about 61%, and at 18 months (using those who could have been followed up within that timeframe as the denominator) it is about 67% (table 21)—rates that are not ideal, but acceptable.
CT PANSS total at 9 months
TAU PANSS total at 9 months
Age (years)
Early intervention services
54·20 (12·35)
63·73 (15·69)
23·45 (6·43)
Other community services
61·08 (16·75)
62·38 (7·39)
41·63 (11·12)
Data are mean (SD). CT=cognitive therapy. TAU=treatment as usual. PANSS=positive and negative syndrome scale.
Table: PANSS score for each group and services setting
401
Wavebreak Media Ltd./Corbis
Correspondence
Published Online June 23, 2014 http://dx.doi.org/10.1016/ S0140-6736(14)61002-3
Published Online June 23, 2014 http://dx.doi.org/10.1016/ S0140-6736(14)61003-5
402
Two correspondents question the use of a mixed (random) effects model based on an assumption that outcome data are missing at random. We provided a technical reference to indicate what was meant by missing at random.3 Its meaning is not quite the same as that implied in everyday English, the latter having the technical description more akin to “missing completely at random” (MCAR). Clearly, MCAR does not hold. The missing-atrandom assumption will only ever be an approximation of the truth (like the modelling assumptions behind any statistical analyses). Our critics seem to assume that if the correct missing data mechanism had been chosen then the estimated treatment effect would be closer to the null. We suggest they would not have been prepared to accept modifications to the prespecified analysis if it led to a greater treatment effect. As we drift away from the prespecified statistical analysis plans, the risk of selecting methods of analysis that favour the outcomes we are looking for becomes a much more serious cause for concern than minor violations of the assumptions underlying the main analyses. Although missing data lead to lower precision (reduced statistical power) and potential biases, the fact that in the present trial we found statistically significant effects implies that it was not underpowered. Lack of power would only have been a serious concern if we had observed a promising but statistically insignificant result. In response to the hypothesis concerning loss to follow-up in generic community services versus early intervention services, we had 9-month PANSS outcomes for ten of 20 (50%) and 12 of 17 (71%) patients allocated to cognitive therapy in early intervention services and community services, respectively. The numbers for treatment as usual (TAU) were 15 of 22 (68%) and 8 of 15 (53%). In a comparison of the proportions
of missing data in the four groups, χ2=2·52 (df=3, p=0·472). We check the sensitivity of our results to differences in proportions of missing data by adding service type as an extra covariate in the mixed-effects model (again assuming missing at random). The revised estimate of the treatment effect on PANSS total is –6·64 (standard error [SE] 2·17), the original estimate being –6·52 (SE 2·18). With the addition of a treatment by service interaction to the mixed model, the effect of cognitive therapy in early intervention services is estimated to be –10·97 (SE 2·78; p<0·001) and the interaction is estimated to be +9·92 (SE 4·22; p=0·019), implying that there is little or no cognitive therapy effect in community services participants (–10·97+9.92=–1·05). This finding that people in early intervention services seem to respond better to cognitive therapy (if real) might not have arisen through differences in the type of service per se, but could be explained by age differences (table). Analysis of the confounding effects of medication would be overly complex for a pilot trial, given the varying timepoints at which it commenced and, more importantly, changes in medication are themselves an outcome of the treatment (ie, potential mediators rather than confounders). We recognise that medication might have had an effect, but with equal proportions in both groups it cannot account for the between-group differences observed. We declare no other competing interests apart from those declared in the original Article.1
*Anthony P Morrison, Graham Dunn, Douglas Turkington, Melissa Pyle, Paul Hutton [email protected] School of Psychological Sciences, University of Manchester, Manchester M13 9PL, UK (APM, MP, PH); Greater Manchester West Mental Health NHS Foundation Trust, Manchester, UK (APM, MP, PH); Centre for Biostatistics, University of Manchester, Manchester, UK (GD); and Newcastle University and Northumberland, Tyne and Wear NHS Mental Health Foundation Trust, Newcastle-upon-Tyne, UK (DT)
1
2
3
Morrison AP, Turkington D, Pyle M, et al. Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial. Lancet 2014; 383: 1395–403. Morrison AP, Wardle M, Hutton P, et al. Assessing cognitive therapy instead of neuroleptics: rationale, study design and sample characteristics of the ACTION trial. Psychosis 2013; 5: 82–92. Little RJA, Rubin DB. Statistical analysis with missing data. London: John Wiley and Sons, 2002.
Department of Error Dickson KE, Simen-Kapeu A, Kinney MV, et al, for The Lancet Every Newborn Study Group. Health-systems bottlenecks and strategies to accelerate scale-up in countries. Lancet 2014; 384: 438–454—In this Series paper, the title should have read “Every Newborn: health-systems bottlenecks and strategies to accelerate scale-up in countries”. Also, the mortality goal listed in the seventh sentence of the Summary should have read “ten or fewer per 1000 livebirths”. Additionally, in the Introduction, the first sentence of the third paragraph should have read “…universal coverage of maternal and newborn care would avert 54% of maternal deaths, 71% of newborn deaths, and 33% of stillbirths…”. These changes have been made to the online version as of June 23, 2014, and to the printed paper. Mason E, McDougall L, Lawn JE, et al, for The Lancet Every Newborn Study Group, on behalf of the Every Newborn Steering Committee. From evidence to action to deliver a healthy start for the next generation. Lancet 2014; 384: 455–67—In this Series paper, the 2035 mortality goals should have read “ten or fewer per 1000” in the third sentence of the Summary, in sentence four of the “Assessment of progress and definition of priorities for action” section, panels 1 and 3, and figure 4. The 2030 mortality goals should have read “12 or fewer per 1000” in sentence four of the “Assessment of progress and definition of priorities for action” section, panel 3, and figure 4. Additionally, in figure 4, the 2020 and 2025 coverage goals should have been preceded by “>”. Further changes have been made to figure 3. These corrections have been made to the online version as of June 23, 2014, and to the printed paper.
www.thelancet.com Vol 384 August 2, 2014
Anthony Morrison and colleagues1 provide an interesting perspective on the use of cognitive behavioural therapy (CBT) in patients with schizophrenia spectrum disorders through their pilot study. Although compliance to CBT was reportedly measured, the extent of the adherence to the theoretical construct in actual delivery was not reported in the paper. 1 Although competence of staff providing CBT was deemed adequate, whether and to what degree the therapists used CBT principles during the sessions was not clear. Since this pilot study focused on feasibility, the measurement of such fidelity to psychotherapeutic construct would be important to ascribe the benefits of CBT to the therapeutic process. CBT for patients with schizophrenia spectrum disorders is quite difficult and requires modification of the usual techniques. 2 Engagement is often considered difficult in such patients, so a retention rate of more than 80% by sixth session is quite encouraging. The mean number of sessions (13·3 sessions) suggests that most participants did not complete the planned 26 sessions. Other medication effectiveness studies like CATIE and CUtLASS3,4 also reported similar substantial dropout rates. The authors believe that about 20% of the participants subsequently started antipsychotic drugs. What is not clear from the paper is the proportion of participants in the cognitive therapy group who were prescribed antipsychotics: 10 of 37 would be about 27%. Some patients who had refused medications were subsequently given medications, suggesting that refusal is not definitive. Alternatively, cognitive behavioural measures might make a person think more positively about medications. This study 1 provides valuable insights about a potential treatment where therapeutic alliance might be difficult and drug side-effects might www.thelancet.com Vol 384 August 2, 2014
impair such an alliance. This approach paves the way for more research into other acceptable yet potentially effective alternative treatment modalities for the treatment of schizophrenia in patients who refuse medications.5 We declare no competing interests.
*Sagar Chandra Bera, Siddharth Sarkar [email protected] Department of Psychiatry, AIIMS, New Delhi, India (SCB); and Department of Psychiatry, JIPMER, Puducherry, India (SS) 1
2
3
4
5
Morrison AP, Turkington D, Pyle M, et al. Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial. Lancet 2014; 383: 1395–403. Tai S, Turkington D. The evolution of cognitive behavior therapy for schizophrenia: current practice and recent developments. Schizophr Bull 2009; 35: 865–73. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med 2005; 353: 1209–23. Jones PB, Barnes TRE, Davies L, et al. Randomized controlled trial of the effect on Quality of Life of second- vs first-generation antipsychotic drugs in schizophrenia: Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS 1). Arch Gen Psychiatry 2006; 63: 1079–87. Lee MS, Shin B-C, Ronan P, Ernst E. Acupuncture for schizophrenia: a systematic review and meta-analysis. Int J Clin Pract 2009; 63: 1622–33.
Authors’ reply Our trial1 was not designed to change clinical practice. It was a preliminary trial, which needs to be followed up by a larger, pragmatic multicentre study. It is important not to overinterpret our data, and we explicitly advised against discontinuation of medication. We claimed the trial showed that cognitive therapy was safe and acceptable, not safe and effective. No participants were included on the basis of attenuated
psychotic symptoms (our entry criteria included positive symptoms above a threshold on the positive and negative syndrome scale [PANSS]). Our combined sample had a mean baseline PANSS score of 72, with the recommended cutoffs for mildly ill being 58 and moderately ill being 75; therefore, our sample consists of people with both levels of difficulty. Duration of psychosis and other sample characteristics are reported elsewhere.2 We can clarify that the proportion of patients starting medication in each group was 27%. An investigation of the components of therapy is being undertaken and will be published. Standardised effect sizes (and 95% CIs) for the PANSS subscales are as follows: PANSS positive (–0·45, –0·81 to –0·09), PANSS negative (–0·22, –0·51 to 0·07), and PANSS general (–0·47, –0·78 to –0·16). The claim that missing data is over 50% at follow-up points is based on use of the randomised number as denominator, rather than the number who could have potentially provided outcome assessments given the planned variable length of follow-up (adopted to provide value for money to the funder, allowing maximal recruitment, and the most complete data we could gather in the lifetime of the trial). The proportion of non-missing PANSS data at 9 months (the final point at which an outcome assessment could have been provided by all randomised participants) is about 61%, and at 18 months (using those who could have been followed up within that timeframe as the denominator) it is about 67% (table 21)—rates that are not ideal, but acceptable.
CT PANSS total at 9 months
TAU PANSS total at 9 months
Age (years)
Early intervention services
54·20 (12·35)
63·73 (15·69)
23·45 (6·43)
Other community services
61·08 (16·75)
62·38 (7·39)
41·63 (11·12)
Data are mean (SD). CT=cognitive therapy. TAU=treatment as usual. PANSS=positive and negative syndrome scale.
Table: PANSS score for each group and services setting
401
Wavebreak Media Ltd./Corbis
Correspondence
Published Online June 23, 2014 http://dx.doi.org/10.1016/ S0140-6736(14)61002-3
Published Online June 23, 2014 http://dx.doi.org/10.1016/ S0140-6736(14)61003-5
402
Two correspondents question the use of a mixed (random) effects model based on an assumption that outcome data are missing at random. We provided a technical reference to indicate what was meant by missing at random.3 Its meaning is not quite the same as that implied in everyday English, the latter having the technical description more akin to “missing completely at random” (MCAR). Clearly, MCAR does not hold. The missing-atrandom assumption will only ever be an approximation of the truth (like the modelling assumptions behind any statistical analyses). Our critics seem to assume that if the correct missing data mechanism had been chosen then the estimated treatment effect would be closer to the null. We suggest they would not have been prepared to accept modifications to the prespecified analysis if it led to a greater treatment effect. As we drift away from the prespecified statistical analysis plans, the risk of selecting methods of analysis that favour the outcomes we are looking for becomes a much more serious cause for concern than minor violations of the assumptions underlying the main analyses. Although missing data lead to lower precision (reduced statistical power) and potential biases, the fact that in the present trial we found statistically significant effects implies that it was not underpowered. Lack of power would only have been a serious concern if we had observed a promising but statistically insignificant result. In response to the hypothesis concerning loss to follow-up in generic community services versus early intervention services, we had 9-month PANSS outcomes for ten of 20 (50%) and 12 of 17 (71%) patients allocated to cognitive therapy in early intervention services and community services, respectively. The numbers for treatment as usual (TAU) were 15 of 22 (68%) and 8 of 15 (53%). In a comparison of the proportions
of missing data in the four groups, χ2=2·52 (df=3, p=0·472). We check the sensitivity of our results to differences in proportions of missing data by adding service type as an extra covariate in the mixed-effects model (again assuming missing at random). The revised estimate of the treatment effect on PANSS total is –6·64 (standard error [SE] 2·17), the original estimate being –6·52 (SE 2·18). With the addition of a treatment by service interaction to the mixed model, the effect of cognitive therapy in early intervention services is estimated to be –10·97 (SE 2·78; p<0·001) and the interaction is estimated to be +9·92 (SE 4·22; p=0·019), implying that there is little or no cognitive therapy effect in community services participants (–10·97+9.92=–1·05). This finding that people in early intervention services seem to respond better to cognitive therapy (if real) might not have arisen through differences in the type of service per se, but could be explained by age differences (table). Analysis of the confounding effects of medication would be overly complex for a pilot trial, given the varying timepoints at which it commenced and, more importantly, changes in medication are themselves an outcome of the treatment (ie, potential mediators rather than confounders). We recognise that medication might have had an effect, but with equal proportions in both groups it cannot account for the between-group differences observed. We declare no other competing interests apart from those declared in the original Article.1
*Anthony P Morrison, Graham Dunn, Douglas Turkington, Melissa Pyle, Paul Hutton [email protected] School of Psychological Sciences, University of Manchester, Manchester M13 9PL, UK (APM, MP, PH); Greater Manchester West Mental Health NHS Foundation Trust, Manchester, UK (APM, MP, PH); Centre for Biostatistics, University of Manchester, Manchester, UK (GD); and Newcastle University and Northumberland, Tyne and Wear NHS Mental Health Foundation Trust, Newcastle-upon-Tyne, UK (DT)
1
2
3
Morrison AP, Turkington D, Pyle M, et al. Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial. Lancet 2014; 383: 1395–403. Morrison AP, Wardle M, Hutton P, et al. Assessing cognitive therapy instead of neuroleptics: rationale, study design and sample characteristics of the ACTION trial. Psychosis 2013; 5: 82–92. Little RJA, Rubin DB. Statistical analysis with missing data. London: John Wiley and Sons, 2002.
Department of Error Dickson KE, Simen-Kapeu A, Kinney MV, et al, for The Lancet Every Newborn Study Group. Health-systems bottlenecks and strategies to accelerate scale-up in countries. Lancet 2014; 384: 438–454—In this Series paper, the title should have read “Every Newborn: health-systems bottlenecks and strategies to accelerate scale-up in countries”. Also, the mortality goal listed in the seventh sentence of the Summary should have read “ten or fewer per 1000 livebirths”. Additionally, in the Introduction, the first sentence of the third paragraph should have read “…universal coverage of maternal and newborn care would avert 54% of maternal deaths, 71% of newborn deaths, and 33% of stillbirths…”. These changes have been made to the online version as of June 23, 2014, and to the printed paper. Mason E, McDougall L, Lawn JE, et al, for The Lancet Every Newborn Study Group, on behalf of the Every Newborn Steering Committee. From evidence to action to deliver a healthy start for the next generation. Lancet 2014; 384: 455–67—In this Series paper, the 2035 mortality goals should have read “ten or fewer per 1000” in the third sentence of the Summary, in sentence four of the “Assessment of progress and definition of priorities for action” section, panels 1 and 3, and figure 4. The 2030 mortality goals should have read “12 or fewer per 1000” in sentence four of the “Assessment of progress and definition of priorities for action” section, panel 3, and figure 4. Additionally, in figure 4, the 2020 and 2025 coverage goals should have been preceded by “>”. Further changes have been made to figure 3. These corrections have been made to the online version as of June 23, 2014, and to the printed paper.
www.thelancet.com Vol 384 August 2, 2014