Now that prospective payment has imposed limits on consumption oflaboratory resources, inappropriate test utilization by clinicians canrepresent a serious dollar drain. Microbiology, one of thelaboratory’s least automated and most labor-intensive areas, isparticularly liable to incur financial losses from excessive testordering. The main problem laboratories face is inappropriate utilization byclinicians. In the not too distant past, underutilization of themicrobiology laboratory was common. A recent requirement by the JointCommission on Accreditation of Hospitals for some form of audit ofantibiotic usage has led most hospitals to require that appropriatecultures be taken, so underutilization is becoming less of a problem. Clinical laboratory services are very susceptible tooverutilization, however. Since hospitals will no longer be reimbursedfor the cost of each laboratory test under prospective payment,excessive laboratory usage can very easily result in a significant lossof funds in many patient admissions.
An example of overuse is the ordering of differential counts.Shapiro et al analyzed the use of differential counts in a 700-beduniversity teaching hospital and found that 26 per cent of alldifferentials done in the hospital laboratory were unjustified.Forty-eight per cent came from medical services, 62 per cent fromsurgical services.
The test results affected patient management in lessthan 3 per cent of patients, and no unjustified tests altered apatient’s diagnosis or treatment. Elimination of unjustifieddifferentials would permit a reduction in 1.8 FTEs from their hospitallaboratory.
It was concluded that the differential is overused andamenable to real cost reduction. Because of the lack of automation presently available to theclinical microbiology laboratory, there is a direct relationship betweenthe workload and the number of technologists needed. Technologists’salaries account for 50 to 70 per cent of a laboratory’smicrobiology is an expensive laboratory service.
It has been noted thatrepetitive, often daily cultures from infected foci clearly reflectconcern on the part of the physician but contribute little or nothing topatient care. Examples of this overuse include excessive numbers ofblood cultures, daily sputum cultures from tracheostomized patientswithout evidence of pneumonia, daily cultures of drainage from infectedfoci, and Foley catheter tip cultures. A study at one hospital revealed that the average patient had 6.1bacteriology tests per admission.
Some patients had as many as 122bacteriology tests, including up to 24 urine cultures. A typicalpulmonary patient in this study had an average of 16 sputum cultures, 13blood cultures, and 9 urine cultures! To promote the optimal use of laboratory tests, we must considernot only factors responsible for inappropriate or excessive use, butalso those that foster underuse. The latter include failure to reviewtest results and inability to interpret them. Optimizing laboratoryutilization requires explicit criteria regarding when laboratory testsshould be used and development of methods to insure that the resultingdata are utilized properly. To be appropriate, a test should affect a patient managementdecision. In microbiology, this primarily occurs through demonstrationand identification of microorganisms and determination of theirantibiotic susceptibilities. The process usually involves genus identification andsusceptibility testing and/or precise identification to species orsubspecies level. For most physicians, the former has a greaterpriority than precise identification.
Besides providing information clinical requirements. Examples ofthis would include the typing of Haemophilus influenzae other than b,typing of Neisseria meningitidis and Klebsiella pneumoniae, and thespecies identification of yeast from mixed cultures. Specimens of dubious value such as those from mouth lesions, bowelcontents, perirectal abscesses, decubiti, vaginal discharges, and Foleycatheter tips should not be processed.
It is not clinically helpful toprovide complete identification and antibiotic susceptibilities of morethan three pathogens in a specimen. Recovery of three or more pathogensgenerally reflects contamination of the specimen with indigenous flora. Concerned over the possible misuse of laboratory services, thestate of Connecticut in 1975 formed an ad hoc committee of clinicalmicrobiologists and pathologists to investigate the utilization ofmicrobiology laboratories.
Two years later, the College of AmericanPathologists convened a meeting in which infectious disease specialists,pediatricians, surgeons, and other physicians met with clinicalmicrobiologists and pathologists to consider clinical relevance inmicrobiology. The findings and recommendations of both meetings were strikinglysimilar. There was a recommendation that institutions establish apolicy for controlling laboratory utilization. Cost containment measures in medical practice will not besuccessful if they are initiated by administrative fiat. Conscientiousphysicians will rebel against constraints they see as deleterious topatient care. Demonstrating what is useful, as opposed to what isunnecessary or harmful, is the cornerstone of scientific medicine.While it is desirable that physicians know the sensitivity, specificity,and predictive values of laboratory tests, they cannot know theperformance characteristics of all the tests they order. To minimizeinefficient and ineffective practice, it is essential that cliniciansand laboratory directors develop guidelines for testing.
The laboratory director should submit specific recommendations tothe directors of the clinical services for their input. When agreementhas been reached, the decisions should be approved by the medical boardand incorporated into hospital policy by the administration. Allattending physicians and the house staff should then be advised of thenew laboratory policies and assured that appropriate exceptions can bemade by consulting with the laboratory director. The activeparticipation of the chiefs of service through chart review is alsoimportant for the success of cost containment programs. Clinical laboratories can contribute further to improving the useof tests.
Laboratory request forms need to be examined. Forms thatlist the entire inventory of available tests are an open invitation tothe user to check off excessive numbers of tests. Active interventionto reinforce an educational program may be required to overcomeestablished patterns of misuse of laboratory resources. Specific strategies for cost containment include the following: 1. Have as much testing done as possible on an outpatient basis,preadmission and post-discharge. This reduces length of stay andremoves some testing from prospective payment limits. 2. Revise laboratory request forms that list tests or procedures bycheck-off boxes.
Provide a separate menu that lists specific tests.For example, when submitting female genital specimens, physicians shouldstate whether they are for possible gonorrhea, vaginitis, surgical woundinfection, or possible endometritis. 3. Attempt to shorten turnaround time through faster ways ofgetting the specimen to the laboratory, testing it, and delivering theresults. Stat tests are disastrous from the perspective of laboratoryefficiency and cost control, but are often necessary from a medicalstandpoint. Many Stats can be avoided, however, by better routine schedulingand establishing priority of tests based upon clinical urgency.
Aprocedure assigning priority to microbiology specimens based upon theirrelevance to the urgency of the clinical problem has been described byEllner et al. Reorganize the laboratory into Stat and routine sectionsto improve productivity. 4. Eliminate tests that are unnecessry–i. e.
, that do notcontribute to the patient’s diagnosis, management, or prognosis.Typical examples of these are throat cultures for organisms other thanbeta hemolytic streptococci; identification of more than three pathogensin a specimen; cultures of oropharyngeal material, bowel contents,perirectal abscesses, decubiti, and Foley catheter tips, and anaerobic culture of surface wounds; and antibiotic susceptibility determinationson obvious contaminants, commensal species from sites they normallyinhabit, or species with predictable susceptibilities. 5. Curb ordering of esoteric tests, and require the prior approvalof an infectious disease specialist. An example is the culture ofcerebrospinal fluid for acid-fast bacilli. 6.
Eliminate replicate and redundant specimens by computermonitoring at registration. The laboratory cannot afford to performthree cultures on the same wound because an intern, a resident, and theattending physician all ordered one independently on the same day. 7. Eliminate tests done by standing orders or routine protocol.Typical examples are tracheal suction specimens from patients in theintensive care unit with no clinical evidence of pneumonia, andperitoneal dialysis specimens from patients with no signs ofperitonitis. 8.
Limit the maximum number of specimens of each type that will beaccepted from a patient. Some suggestions are shown in Figure I. 9. Develop progressive test profiling, in which follow-up tests areautomatically performed according to predetermined, cost-efficientalgorithms designed to obtain the most diagnostic information. Examplesare the testing of serum for syphilis, hepatitis, or streptococcal antibodies. The laboratory can contribute to the more cost-effective use ofantibiotics by selective reporting. In our institution,second-generation cephalosporin results are reported only if theorganism is resistant to cefazolin, and third-generation drugs arereported only if the isolate is resistant to both cefoxitin andcefamandole. 10.
Eliminate routine environmental cultures. These includesamples from floors, surfaces, air, presterilized fluids, medications,or infant formula milk. 11.
Cut total costs by sharing lab services and resources withother hospitals. Economies of scale result from this approach. 12. Increase test volume from sources outside the hospital tocompensate for cuts in inpatient volume, and adopt competitiveoutpatient pricing. Fixed laboratory costs remain the same. The addedbusiness will entail only variable-cost increases.