Article
Plug-and-play-mentality vs. quality management in ICP-CPP-measurement
Search Medline for
Authors
-
A. Aschoff
- Neurochirurgische Klinik, Universitätsklinikum Heidelberg
| Published: | May 20, 2009 |
|---|
Outline
Text
Objective: Between 1960 and 1980 on congresses and in journals related to neurointensive care the technical and methodological aspects of ICP-measurement (ICP-M) played a significant role. Sensor drift, linearity failures and even technical details of transducers were common topics, and critical attitudes concerning the detection of artifacts, methodological problems and the reliability of ICP-M were the consequence.
In the mid-nineties the intraparenchymatous transducers and other progress suggested the end of measurement problems (“Plug and play”). The number of methodological papers collapsed, the young generation lost know-how.
Methods: Based on personal experience with ventricular, epidural, subdural, and parenchymatous measurements using 15 different transducers in >3,000 patients since 1977, and on lab tests of >40 probes we demonstrate typical artifacts and failures.
Results: Drift and accuracy problems common in the 80s decreased, but did not disappear. Commonly used transducers sometimes fail by 10 mmHg and more. A post-explantation study of Gelabert-González (2006) reported a rate of only 12.3% accurate (= failure ±<2 mmHg) transducers (624 probes). Every second transducer of a leading producer failed even >5–20 mmHg! Between calibration temperatures of 20oC vs. brain (38°C) most transducers differ by 3–5 mmHg. In the clinical practice fluid coupling problems are omnipresent and difficult to control, especially in automatic, computerized registrations. There is a jungle of wave-form deviations due to damping, resonance, time-shifts and other effects; in addition there are frequent level discrepancies between transducers vs. head or heart. The “gold standard“ quality of ventricular ICP registration is a myth.
Conclusions: We must interpret ICP/CPP data with permanent sensitivity for potential artifacts. All transducers need a final check after explantation and a short documentation of problems during implantations (“events”). The impact of arterial pressure measurement failures on CPP is underestimated. We warn in particular against automated or computerized data registration without regular controls and redundancy.
