Methodologic issues associated with secretion weight as a dependent variable in research using closed-system suction catheters

Background: Closed-system suction catheters (CSSC) were designed to elimina te the need to disconnect the patient from the ventilator during endotrache al suctioning (ETS). During data collection on an NIH-funded study, it was noted that moisture accumulated on the inside of the CSSC and sleeve when a ttached to the patient for 30 minutes. Because CSSC are not disconnected, t hey present unique methodologic problems related to measurement of secretio ns as a dependent variable in clinical research. Objectives: To describe a valid, reliable, and practical method for weighin g secretions obtained during ETS using a CSSC; and to determine the change in weight of a CSSC after its attachment for 30 minutes to a mechanically v entilated patient. Methods: After being weighed, a CSSC and sputum trap were attached to the e ndotracheal tube of a mechanically ventilated adult and remained attached f or 43 minutes (30 minutes to allow positive end expiratory pressure and oxy genation levels to return to normal and 13 minutes to mimic the time frame for the ETS procedure used in another study). No ETS occurred. The CSSC and sputum trap were then removed and reweighed. Results: A convenience sample consisted of 50 adults who were critically il l and mechanically ventilated. Independent variables included tidal volume, pressure support, body temperature, and respiratory rate. The dependent va riable was wet weight of the CSSC, determined by subtracting the preprotoco l catheter weight from the postprotocol catheter weight. The mean wet weigh t for all catheters was 0.5142 +/- 0.1215 grams. In a subset of 37 patients , two wet weights (74 paired observations) were determined. The mean wet we ight for these catheters was 0.54014 +/- 0.1404 grams. The paired wet weigh ts were statistically different (t = 2.433; df = 36; p = 0.02). Pearson cor relation coefficients and beta coefficients were computed. While tidal volu me and pressure support were highly correlated (r = 0.678; p = 0.011), ther e were no other statistically significant associations. Conclusions: The amount of secretions is a common dependent variable in ETS research. During the time that CSSC are attached to the patient and ventil ator, moisture from either the ventilator's humidification system or the pa tient accumulates in the CSSC. This wet weight is not actually part of the secretions retrieved during ETS and should not be considered in the actual weight of secretions. Further study on the determinants of wet weight is wa rranted.