Letters to the Editor Activated Protein C: More Effective than Nonactivated Protein C? To the Editor: Dr. Dhainaut and colleagues (1) rec- ommend that the activated form of pro- tein C should be preferred over the zy- mogen form of protein C for the treatment of patients with severe sepsis. This suggestion is primarily based on a report by Faust et al. (2), which contains data that are inconsistent. Faust et al. state that the administration of 50 IU/kg of protein C every 8 hrs (on days 1, 2, and 3) resulted in normal trough levels of protein C antigen (as measured immedi- ately before an infusion) and peak levels (measured 1 hr after an infusion). It is highly unlikely that a single dose of 50 IU/kg would result in normal levels of protein C antigen being sustained for up to 8 hrs in a patient with severe sepsis in whom coagulation is activated and pro- tein C consumed (3). Furthermore, based on the pharmacokinetic properties of protein C, it is impossible that its infu- sion into a patient with a normal protein C level would result in a peak that is within the normal range, rather it would be far above the normal range. The assay used by Faust et al. mea- sures only the free activated protein C (APC) levels in a plasma sample at the time of drawing blood. In vivo-generated APC forms complexes with serpins such as -2 macroglobulin, -1 antitrypsin, protein C inhibitor, plasminogen activa- tor inhibitor, and also ceruloplasmin. Measurement of free APC with the assay used does not, therefore, reflect the total but an underestimation of the APC gen- erated in vivo. Faust et al. immunohistochemically show a down-regulation of thrombo- modulin in skin biopsy specimens taken from the edge of purpuric or petechial lesions. Conclusions drawn from this semiquantitative assessment of thrombo- modulin expression in affected areas of the skin neglect the mass of endothelial cells in other organs not affected by pur- pura fulminans where normal expression of thrombomodulin exists, as shown in sepsis models (4), providing for normal conversion of protein C zymogen to APC. Treating patients who have purpura fulminans and meningococcal sepsis with protein C zymogen halts the pro- gression of skin lesions, suggesting that protein C zymogen can act locally, even in an area of skin where Faust et al. found semiquantitative down-regulation of thrombomodulin and endothelial pro- tein C receptor. In view of these considerations, and in the absence of clinical data of a com- parative study, it is inappropriate to conclude that APC may be more effec- tive than nonactivated protein C in these patients. Hans Peter Schwarz, MD, Hartmut J. Ehrlich, MD, Baxter BioScience, Vienna, Austria REFERENCES 1. Dhainaut J-F, Yan SB, Cariou A, et al: Soluble thrombomodulin, plasma-derived unactivated protein C, and recombinant human activated protein C in sepsis. Crit Care Med 2002; 30(Suppl):S318 –S324 2. Faust SN, Levin M, Harrison OB, et al: Dys- function of endothelial protein C activation in severe meningococcal sepsis. N Engl J Med 2001; 345:408 – 416 3. Smith OP, White B, Vaughan D, et al: Use of protein-c concentrate, herapin, and haemodi- afiltration in meningococcus-induced purpura fulminans. Lancet 1997; 350:1590 –1593 4. Laszik Z, Carson CW, Nadasdy T, et al: Lack of suppressed renal thrombomodulin expression in a septic rat model with glomerular throm- botic microangiopathy. Lab Invest 1994; 70:862– 867 DOI: 10.1097/01.CCM.0000065765.58732.C3 Family Satisfaction with Intensive Care Unit Care: Influenced by Workload, Staffing, and Patient Selection? To the Editor: We read with interest the article by Dr. Heyland and colleagues (1) in the July issue of Critical Care Medicine. The au- thors studied levels of family satisfaction with critical care performance in six in- tensive care units (ICUs) across Canada, using a questionnaire given to a patient’s family members when the patient left the ICU. If a patient died in the ICU, the questionnaire was sent to family mem- bers some weeks after the patient’s death. The authors report that levels of satisfac- tion with ICU care were high, with 84.3% of respondents being satisfied or highly satisfied with overall ICU care. We compliment the authors on ad- dressing this important topic and for the thorough way in which they approached this issue. We would, however, like to make two comments. The authors report that the majority of respondents were sat- isfied with overall care in the ICU, judg- ing this to be excellent or very good. The most important determinant appeared to be the quality and quantity of nursing and physician communication with fam- ily members (1). In our opinion, this raises an important question that was not addressed in this study: namely, whether there was a relationship between work- load and staffing levels in each ICU and the degree of family satisfaction. A num- ber of studies suggest that overall quality of care and outcome in ICUs may be linked to workload and staffing levels (2– 4); therefore, in our opinion, it is plausi- ble that patient and family satisfaction may also be influenced by these factors. The study by Dr. Heyland and colleagues was a multiple centered study, and staff- ing levels within individual ICUs are likely to have varied over time. Therefore, it should be possible to address the issue of whether the degree of family satisfac- tion was influenced by the number of nurses and physicians per patient and by their workload. Indeed, perhaps this could be addressed from the data already collected by the authors. Perhaps they could comment on this subject. Second, the authors did not directly address the issue of potential bias, for example, that a disproportionate number of nonresponders to the questionnaire may have been less satisfied with the overall levels of ICU care than respond- ers. Although this is to some degree in- evitable, and the response rate in this study (70%) was relatively high, we won- der how the authors tried to avoid this problem. For example, it is stated in the Method section that a research nurse pro- spectively approached “eligible” patients to solicit their involvement in the study. Does this mean that every patient was Copyright © 2003 by Lippincott Williams & Wilkins 1597 Crit Care Med 2003 Vol. 31, No. 5