Our learning disability team recently started clozapine for a young patient with treatment-resistant schizophrenia and a moderate intellectual disability. As treatment progressed, improvements were noted in the patient's positive symptoms and, perhaps more importantly, they started gaining some insight. However, although the patient credited the clozapine for this improvement, they were increasingly against the idea of regular blood monitoring. Continuing this monitoring in the community without their cooperation was not going to be possible. However, the team wanted the patient to continue to have the benefits the clozapine treatment was affording them.

Although venous blood sampling is the preferred method for monitoring neutrophil counts in adults, the team wondered whether our patient would be more amenable to blood monitoring if we used a capillary sampling technique favoured in paediatrics. There have been a few published papers looking at the variation in results for venous and capillary samples, but some studies had small sample sizes ^{Reference Yang, Yang, Chen, Qu, Zhu and Tang1} and others have included participants far younger than our patient. ^{Reference Daae, Hallerud and Halvorsen2}

However, in a letter to the American Journal of Haematology, Schalk and colleagues ^{Reference Schalk, Scheinpflug and Mohren3} describe a study they undertook involving 421 adult patients, 70% of whom had a haematological disorder and 30% of whom were healthy volunteers; the age range was 18–61 years. They concluded that capillary and venous absolute neutrophil counts correlate very well in adults. They also found there were no higher rates of infection using capillary sampling compared with venous sampling for their patients with neutropenia or agranulocytosis. They note that previous studies showed capillary samples had higher absolute neutrophil counts than venous samples but that this seemed to decrease with the increasing age of the cohort. In the 3 months to 14 years age group the variation between venous and capillary samples was 17.2%, whereas in the 20–22 years age group the difference in total leucocyte count was 9.2%. ^{Reference Yang, Yang, Chen, Qu, Zhu and Tang1,Reference Daae, Hallerud and Halvorsen2} Both these studies are limited by small sample sizes (n = 9 and n = 24 respectively).

The hospital pharmacy and clozapine monitoring service agreed for us to proceed with capillary sampling for our patient. The paediatric nurses provided practical advice for taking the sample. This included using petroleum jelly around the finger to make the blood form droplets as well as ‘milking’ the arm to reduce discomfort.

Although this method took more time than venous sampling, the patient was very happy with the finger-prick technique and volunteered to have a sample done with no extra effort. They were able to have this done in the community and were discharged.

Although venous sampling remains the method of choice for monitoring patients on clozapine, we feel capillary sampling could be an option for patients who are unwilling to have the venous injection or where venous access proves difficult.