Retroperitoneal space abscesses are unusual clinical problems encountered by general surgeons, internists, and surgical subspecialists. An insidious, occult illness marked by diagnostic delay, inadequate drainage, and considerable morbidity and mortality is common. Anatomic reviews detailing the complex extraperitoneal spaces have been published, but less attention has been focused on diagnostic and drainage techniques useful to the practicing surgeon. In a retrospective review of 50 extraperitoneal abscesses, attention was directed to clinical presentation, diagnosis, and therapy. On the average, 12.7 days were required to establish the diagnosis; 50% of patients suffered major complications. A strikingly high mortality was associated with positive blood cultures and persistent fever within 48 hours of drainage (75% and 71%, respectively). Computed tomography has greatly enhanced the diagnosis of extraperitoneal abscesses, and radiologic drainage in selected cases appears to be a useful initial approach. A simplified anatomic classification and treatment plan is proposed to facilitate comparison between clinical series.

No abstract

Problems encountered in the process of modifying simple operant behavior of a retarded S from what is observed at the beginning of a study to that required by a multiple schedule have two major implications. One bears on an experimental analysis of individual differences; the other, on the development of techniques for the efficient establishment of complex-schedule performances.When a child enters an experimental situation, receives instructions, and sets about to perform the task, the behavior displayed is, of course, a function of the current situation and interactions with similar situations in the S's history. The influences of such antecedents may be conceptualized as effects of independent variables (e.g., kinds of reinforcers received, typical schedules, and frequency of punishment contingencies), and of differences in behavioral processes (e.g., rate of change in operant conditioning) (Skinner, 1953). Such effects may be quantified by psychometric devices suchA as inventories of traits and abilities, or by experimental procedures. The latter, which involve observation of the successive changes in behavior required to perform an experimental task to criterion, may be approached in two ways. One consists of presenting the task and recording time (and "errors"). This procedure is often abortive. If the task is complex, even slightly so, learning may take an unreasonable length of time, or may not be achieved at all. The other approach involves presenting S with a series of graded tasks and reinforcing responses that approximate more and more the final performance required. The procedure is designed so that the S sets the pace; that is, each response class is strengthened to