Gram staining provides a presumptive diagnosis and some indication of the organism involved without waiting for culture results. On occasion, Gram stain may reveal organisms that (for technical reasons) do not grow when cultured. Best information comes from areas that are normally sterile. Gram staining is considered a routine procedure for CSF in possible meningitis, for urethral smears in possible venereal disease, and for material from abscesses or effusions. (This is especially true when anaerobes may be present. These may fail to grow, since “anaerobic” culture is often suboptimal, but the organisms might be seen on a Gram stain.) In certain other types of specimens, such as urine or stool, a Gram stain need not be done routinely but should be performed in special circumstances, such as on stool specimens when pseudomembranous enterocolitis is suspected or on urine specimens if quantitative culture is not available.

The Gram stain is controversial in several other areas. The most important example is sputum. One can find opinions in the literature on the usefulness of a sputum Gram stain ranging from “essential” to “worthless and misleading.” One major problem is contamination by normal organisms from the mouth and nasopharynx. Certain pathogens, such as pneumococci or S. aureus, may be located in the upper respiratory tract of many clinically normal persons, and these may gain entry into the sputum. Similarly, upper respiratory tract colonization by enteric gramegative organisms is relatively frequent in alcoholics and in patients who have been hospitalized for more than a few days. Another difficulty is occasional discovery of opportunist organisms in the lower respiratory tract, organisms not normally present but not apparently causing disease. Another problem is the morphologic similarity of pneumococci and streptococci, among others. Finally, and most important, the sputum sample may, in fact, be only saliva.

There are certain maneuvers that may help to circumvent some of these drawbacks. The number of organisms may provide a clue to their significance. Marked predominance or heavy growth of one organism suggests true abnormality. Many polymorphonuclear leukocytes is some evidence in favor of acute infection. Significant quantities of squamous epithelial cells (>10 per low-power field) usually means oral contamination. The patient should be instructed how to produce a deep cough and, if possible, should be observed while he or she produces the specimen. In some cases aerosol therapy may be helpful to induce sputum.