Intensive care is a major cost component in modern healthcare systems . In our sample, initiation of mechanical ventilation led to a 59% average cost increase, a very considerable increase. While costs of a ventilated ICU day differed very little between the different patient groups, the large variability of the cost increase associated with initiation of ventilation could open up avenues to effective resource allocation by for example focusing preventative measures, where multiple possible interventions might compete for funding, on the patient groups where avoidance of MV would be associated with the highest savings. Detailed cost data is thus useful to inform policy and optimally allocate limited resources. Our findings contribute towards this.
Overall, our results are in line with the available literature [2,3,4,5]: Dasta et al. (2005), for instance found much higher daily costs of a non-ventilated or ventilated bed day at US hospitals (US$3250 and US$4772, in 2002 values), but the relative cost increase (47%) seems comparable . Other studies reported much lower extra costs of mechanical ventilation: Moran et al. (2004) determined the daily costs of a non-ventilated or ventilated bed day at Australian hospitals (AU$1616 and AU$1911, in 1991 values), corresponding to a relative cost increase of 18% . The only other German study on the topic found costs of a non-ventilated or ventilated bed day of EUR 680 and EUR 946 (in 2003 values), a relative cost increase of 39% .
One surprising finding was that ICU patients with musculoskeletal diseases (as main diagnosis) were, on average, associated with very high daily costs even when not ventilated (€ 1357). This was especially marked when compared to ICU patients with respiratory diseases, who were associated with much lower daily costs when not ventilated (€ 795). This may imply that ICU patients with musculoskeletal diseases, on average, require higher treatment intensity even when they are not ventilated. Among the ICU patients with respiratory diseases (as main diagnosis), in contrast, the absence of ventilation might be associated with a generally lower treatment intensity.
A limitation of our study of course is the single-center nature of the data, however the sample was decently-sized and included all patients treated in the period examined, limiting some sources of bias. The competing interests of the hospital to trigger reimbursement for services rendered and the sickness funds to limit cost should result in a good level of reliability of the administrative data.
Another limitation is that daily ICU costs were not available for analysis. In practice, avoidance of one last additional day of ventilation in a given patient is expected to lead to lower cost savings than avoidance of the most expensive first day of ventilation, after which daily cost drops rapidly . However, this does not detract from the usefulness of our findings on the cost differences between patient groups by the ICD-10 chapter of the main diagnosis, which has been previously underreported and is important due to the large size of the effect.
Key points for decision makers
Mechanical ventilation markedly increases daily ICU cost.
The magnitude of the increase over unventilated care differs strongly between different underlying diseases.
It might be possible to generate saving by focusing budgets for efforts to prevent necessity of ventilation on fields where initiation of ventilation would lead to a particularly pronounced cost increase.