Impact of a peri-operative quality improvement programme on postoperative pulmonary complications

The importance of clinical risk factors for postoperative pulmonary complications and the value of preoperative testing to stratify risk are the subject of debate. To systematically review the literature on preoperative pulmonary risk stratification before noncardiothoracic surgery. MEDLINE search from 1 January 1980 through 30 June 2005 and hand search of the bibliographies of retrieved articles. English-language studies that reported the effect of patient- and procedure-related risk factors and laboratory predictors on postoperative pulmonary complication rates after noncardiothoracic surgery and that met predefined inclusion criteria. The authors used standardized abstraction instruments to extract data on study characteristics, hierarchy of research design, study quality, risk factors, and laboratory predictors. The authors determined random-effects pooled estimate odds ratios and, when appropriate, trim-and-fill estimates for patient- and procedure-related risk factors from studies that used multivariable analyses. They assigned summary strength of evidence scores for each factor. Good evidence supports patient-related risk factors for postoperative pulmonary complications, including advanced age, American Society of Anesthesiologists class 2 or higher, functional dependence, chronic obstructive pulmonary disease, and congestive heart failure. Good evidence supports procedure-related risk factors for postoperative pulmonary complications, including aortic aneurysm repair, nonresective thoracic surgery, abdominal surgery, neurosurgery, emergency surgery, general anesthesia, head and neck surgery, vascular surgery, and prolonged surgery. Among laboratory predictors, good evidence exists only for serum albumin level less than 30 g/L. Insufficient evidence supports preoperative spirometry as a tool to stratify risk. For certain risk factors and laboratory predictors, the literature provides only unadjusted estimates of risk. Prescreening, variable selection algorithms, and publication bias limited reporting of risk factors among studies using multivariable analysis. Selected clinical and laboratory factors allow risk stratification for postoperative pulmonary complications after noncardiothoracic surgery.

No externally validated risk score for postoperative pulmonary complications (PPCs) is currently available. The authors tested the generalizability of the Assess Respiratory Risk in Surgical Patients in Catalonia risk score for PPCs in a large European cohort (Prospective Evaluation of a RIsk Score for postoperative pulmonary COmPlications in Europe).

The Hawthorne effect in experimental research is the unwanted effect of the experimental operations themselves. Following the Hawthorne studies, various explanations have been proposed to account for rising rates of production. Although in the Relay Assembly Test Room experiment the experimental operations may have produced other extraneous variables, a reexamination based on new and neglected evidence has yielded a new interpretation. The new variable, made more plausible because research in other contexts has shown it to have similar effects, is a combination of information feedback and financial reward. It is an example of the control of behavior by its consequences. Although several approaches may be taken to explain the effects of response-consequence contingencies, I have favored operant conditioning because it seems to account for progressive increases in response rate-the Hawthorne phenomenon. Generalizing from the particular situation at Hawthorne, I would define the Hawthorne effect as the confounding that occurs if experimenters fail to realize how the consequences of subjects' performance affect what subjects do. But the Hawthorne effect need not be viewed solely as a problem in conducting experiments. The phenomenon that created it should be studied in its own right, as Sommer (67) suggested with a different phenomenon in mind. The study of response-consequence contingencies might well be extended to the examination of motivation in industrial workers.