We agree that attempts should be made to continue patients on clozapine if at all possible in order to give them an adequate trial of the drug. The Clozaril Patient Monitoring Service (CPMS) routinely advises that samples should not be taken first thing in the morning for patients with borderline white cell/neutrophil counts to avoid the problem of morning pseudoneutropenia. Taking samples after a brisk walk has also been suggested in such patients.

Although Hummer et al (Reference Hummer, Kurz and Barnas1994) reported transient neutropenia in 22% of 68 patients they defined leucopenia as a white cell count <3.5×10⁹/l and neutropenia as a neutrophil count <2.0×10⁹/l. The mean neutrophil count at the time of the transient neutropenia was 1.78×10⁹/l. In the UK and generally, the cut-off points for leucopenia and neutropenia used in clozapine monitoring are lower (3.0×10⁹/l and 1.5×10⁹/l respectively) and patients with counts higher than this are not required to stop clozapine but are monitored more frequently. The relevance of the findings of the study to all clozapine-treated patients must therefore be considered with this point in mind.

The use of a hydrocortisone test to distinguish between benign and malignant neutropenia is of great interest but findings must be interpreted with caution as the study involved only three patients (Reference Murry and LaurentMurry & Laurent, 2001). Furthermore, the risk of further stressing a compromised bone marrow must be borne in mind with such interventions. Whether it is possible to distinguish between transient neutropenia and the prelude to agranulocytosis in clozapine-treated patients remains to be determined.