The birth of the International Classification of Primary Care (ICPC). Serendipity at the border of Lac Léman.

Twenty years ago, in early spring, we were sitting in the Japanese garden of the WHO Headquarters in Geneva. It was the last day of a week of hard work, together with Sue Meads from the US National Centre for Health Statistics (NCHS), on the Reason for Encounter Classification. Since 1978, WHO had—with fiscal support from the NCHS—invited us several times to come over and prepare a classification representing patients' ('subject-ive') demand for care. This seemed important for health care planning, as an addition to data collected with ICD reflecting 'objective' patients' need from a medical perspective. Internationally, the focus of ICD was shifting from mortality towards morbidity, and thus the reasons to visit a doctor became of more interest. The three of us seemed to be rather well equipped for this exercise: two family physicians (FPs) with experience in family practice morbidity statistics, and a taxonomer who, as a country doctor's daughter, felt equally sceptical about many diagnoses and the utility of the resulting interventions. Sitting there, we discussed our present situation. We had just finished the field trial version of the Reason for Encounter Classification, together with a manual for its use. 1 Support from WHO, NCHS and WONCA would soon facilitate field trials in Australia (Charles Bridges (Cees de Geus and Henk Lamberts), Norway (Bent Bentsen) and the USA (Maurice Wood and Sue Meads). The atmosphere in WHO headquarters was optimistic. The Report Health for all in the year 2000 was on its way, as was ICD-10 as the centre of a 'family of classifications'. It was quietly accepted that FPs would not use ICD-10 (~10 000 classes) as a diagnostic classification, but that the ICD-9-related primary care classification ICHPPC-2 (370 classes) would be succeeded by an ICD-10-related version. 2 The trial version of the Reason for Encounter Classification contained ~700 classes. 1 Included were 200 symptoms , complaints, concerns, fears and psychosocial problems not available in ICD-9 and ICHPPC-2, because of the observation that patients often formulate health problems as symptoms and complaints. In addition, they sometimes formulate their problem as a diagnosis (I'm here for my hypertension), so most diagnostic classes of ICHPPC-2 were also included in the new classification. An essential and really new element was the inclusion of reason for encounter rubrics for patients' requests such as: I would like a prescription, a referral, a blood test; would you please measure my blood …