Data base systems contain valuable information to determine the use of health services, variability in clinical practice, and the health care processes, or to evaluate the impact of health service reforms.1
The progressive generalised use of the health card in our country has given a considerable boost to investigation in this field. The increasing computerisation of the health services will also help to make these types of study more popular. The article by Alberdi et al1 is a clear example of a study centred on system databases. It links the information provided from different sources. It uses the health card census, but by only using this register it does not achieve sufficient information to associate the sociodemographic characteristics and the health of each patient with their use of the services. It is surprising that the health care institutions themselves have such difficulty in integrating data from their own registers, particularly when, as regards the autonomous regions, each citizen has had a unique identifier for such a long time. The number of health care cards is used in every one of the events which are produced in the health care process: to arrange an appointment in the primary care centre, to record clinical data in the computerised clinical history, to collect all their drugs from the pharmacy, to be admitted into hospital, and come out with a discharge report. It should not be so difficult to make this combined information on the citizen available for use in the evaluation and planning of health services, as well as in individual clinical care.
It is interesting that the authors should find a territorial pattern in the use of services.2 As they point out, some demographic, socioeconomic and health characteristics partially explain this association. However, it is possible that other factors not analysed in this study could have a decisive influence. In a country like ours, where patients are captive within a sectorised health system, I fear that a large part of this geographic factor is also determined by the level of health services themselves. Aspects such as geographic accessibility, hours and culture of the specialist centres, the variability in clinical practice itself, or the different referral protocols agreed with the family doctors could help to explain this observation. The authors cannot confirm with conviction, that in their territory, the level of services might not explain the use, when many of the variables which characterise it have not been considered.
It is also important for us to understand the limitations of analysing referrals to specialised care based exclusively on a quantitative perspective. We have to consider that a large number of referrals are associated with a limited ability to restrict, and as a result, with a lower clinical quality of the family doctor. But this simplistic interpretation is very wrong.
A comparative analysis of the specialist referrals in the United Kingdom compared to the 5 health plans of the United States shows that, with the same patient load, the British family doctors refer 13.9% of patients, as opposed to the 30% to 36.8% of the Americans.3
It is easy to think the differences in these health, legal, and cultural models of both countries explains the variation. But we cannot conclude that the British doctors may be more resolute and, as a result, more efficient than the North American health plans. A more detailed analysis showed that the Kaiser Permanente model of the plan of a Californian non-profit organisation was much more efficient than the British National Health Service. Effectively, it had a higher participation of specialist care, but the
health processes were more resolute. The secret of the Californian success consisted of, being a more integrated system, some better managed hospitals, working in a competitive environment and they had invested more in information technology.4
The analysis of system data bases is an interesting investigation source which will allow us to better determine the needs of our patients and the ways of approaching them. The information which helps us can have an immediate practical use in improving our health care processes.
