Issues of knowledge, fears, motivation and potential psychological impact of performing CISC should be addressed prior to deciding on CISC and instructing patients. Follow-up care should be improved to include re-evaluations of skills, discussing adherence, integrating CISC in daily activities and general coping issues.

SUMMARY Anticonvulsant medication was stopped in a prospective study in 116 children with epilepsy who had had no seizures for a period of 2 years. A remission rate of 80O5% was found 5 years after withdrawal. The population studied was unselected, and based on children directly referred by general practitioners to the outpatient department. Among the many variables examined, such as type of epilepsy or seizure, presence of concomitant neurological or intellectual deficit, and epileptiform activity on the EEG, only the age of onset of seizures was significantly and positively correlated with the probability of recurrence after discontinuation of medication. In contrast to other recent studies, it was concluded that there are no reliable predictive factors for withholding from any individual "epileptic" child the benefit of attempts to stop medication after 2 years of seizure freedom.The consensus of opinion is that prognosis in childhood epilepsy after discontinuation of anticonvulsants is more favourable than that in adult epilepsy.' However, when fits have stopped in an individual child, the physician will not find easy answers to such questions as: how long should treatment be continued; and are there (clinical) factors which will reliably predict in this particular child whether or not treatment should be stopped? There is some controversy about the significance of various factors (the type of seizure, age at onset of seizures, and, in particular, the electroencephalogram (EEG)) which have bearing on this decision whether to stop or not to stop. These controversies may be partly explained by the differences in populations studied.' 2 The present series of children, which was unselected and based on a patient population directly referred by general practitioners, has been studied in a prospective fashion. The results, based on a 2 year seizure free interval, indicate that 80% of children do not have a recurrence of seizures within 5 years after stopping anti-

Objective Previous studies in patients with lung cancer examined the association between psychological factors with quality of life (QoL), as well as the association between psychological factors with sociodemographic and medical characteristics. However, knowledge about the impact of combinations of psychological characteristics on QoL is still lacking. Therefore, the current study aimed to identify psychological profiles, covering multiple psychological factors. Additionally, the association between these profiles with QoL and with sociodemographic and medical characteristics was explored. Methods Patients with lung cancer (n = 130, mean age = 68.3 ± 8.6 years; 49% men) completed questionnaires focusing on sociodemographic information, anxiety and depressive symptoms (HADS), coping (COPE-easy), perceived social support (PSSS), and QoL (WHOQOL-BREF). Medical information was extracted from patients' medical records. A step-3 latent profile analysis was performed to identify the psychological profiles. Multinomial logit models were used to explore the medical and sociodemographic correlates of the profiles and the relation with QoL. Results Four psychological profiles were identified as follows: (1) anxious, extensive coping repertoire (33%); (2) depressive, avoidant coping (23%); (3) low emotional symptoms, active/social coping (16%); and (4) low emotional symptoms, limited coping repertoire (29%). QoL in profile 1 (QoL = 6.59) was significantly different from QoL in profile 3 (QoL = 8.11, p = .001) and profile 4 (QoL = 7.40, p = .01). QoL in profile 2 (QoL = 6.43) was significantly different from QoL in profile 3 (QoL = 8.11, p = .003) and profile 4 (QoL = 7.40, p = .02). Regarding QoL, no other significant differences were found. Sociodemographic and medical characteristics were not distinctive for the profiles (all p values > .05). Conclusion Determining psychological profiles of patients with lung cancer in an early stage provides information that may be helpful in aligning care with patients' unique needs, as it will help in more adequately selecting those patients who are in need of psychological screening and/or psychological treatment as compared with determining scores on single psychological factors.

No abstract