A common obstacle for appropriate use of antibiotics is often said to be lack of diagnostic tests that can either verify or rule out a bacterial infection, as uncertainty drives use of antibiotics “just in case”. One estimate, for example, states that universal availability to an “optimal” diagnostic test for acute lower respiratory tract infections could save 404 million unnecessary antibiotic treatments in low and middle income countries. But even if a diagnostic test is available it needs to be used and be relied upon by physicians and patients. Two studies from the USA published during the first half of 2018 illustrate some of these problems.
Study 1: Procalcitonin for diagnosis of lower respiratory tract infections
Procalcitonin is an inflammatory biomarker which increases rapidly during systemic bacterial infection. It can therefore be used to determine whether an infection is caused by bacteria or a virus and is thus expected to be able to decrease unnecessary use of antibiotics by two distinctive pathways:
- Elimination of antibiotic treatment for viral infections.
- Shorted treatment duration in bacterial infections by earlier discontinuation when procalcitonin levels decrease.
According to a review by the Swedish Public Health Authority, previous studies have shown clear reduction of both amount of antibiotic courses and treatment duration without maintained disease prognosis.
In study published in New England Journal of Medicine, the effect of introduction of a Procalcitonin diagnostic test was studied in mainly urban hospitals in USA. All participating hospitals have a track record of high adherence to guidelines. The study enrolled 1656 patients with suspected lower respiratory tract infection but with uncertainty about whether antibiotics should be prescribed. The patients were randomized to either the usual care group without procalcitonin data available for the treating physician or to the procalcitonin group where procalcitonin test data was made available to the treating physician. Test data was made available before making the decision of antibiotic treatment and, if the patient was hospitalized, during treatment.
Somewhat surprisingly, no significant difference was seen on total antibiotic use between the groups during the first 30 days: mean antibiotic days in the procalcitonin group was 4.2, whereas the usual care group had 4.3 mean antibiotic days. No difference was seen in subgroup analysis either. At first glance, it seems that the new diagnostic test has no effect on patient outcomes. But given the earlier results showing a clear effect of procalcitonin diagnostics, one must consider a final parameter: adherence to guidelines. Adherence varied depending on the diagnosis: in acute bronchitis, adherence was 82% overall, whereas in community-acquired pneumonia the adherence dropped to 39%. In addition, there appears to have been some contamination between the study groups. So the lack of effect from introduction of procalcitonin diagnostics could be explained by non-adherence to guidelines – especially in potentially more severe infections.
Read more about diagnostics:
Study 2: Improving antibiotic stewardship in upper respiratory tract infections
Upper respiratory tract infections is a disease group where unnecessary use of antibiotics continues to be rampant, even in high income countries with well-developed health care systems like the USA. They are most commonly caused by any of several viruses, but in approximately 20% of the cases, a bacterial cause may be found or strongly suspected. For such cases, antibiotic treatment is warranted. A study in JAMA investigated how antibiotic stewardship could be improved in upper respiratory tract infections. The study enrolled outpatients older than 6 months and collected medical history, diagnose codes, diagnostic tests and antibiotic prescriptions during the influenza seasons 2013-2014 and 2014-2015.
Of the patients in the study, 41% were prescribed antibiotics and of these, 54% had diagnoses where antibiotics are not indicated or did not meet treatment criteria. Also, 56% of antibiotic prescriptions were on broad-spectrum antibiotics. A subgroup of patients were ones with pharyngitis, an infection where a simple diagnostic test is commonly used to verify the presence of Group A Streptococci, GAS. In the subset of patients with only pharyngitis, 38% were prescribed an antibiotic in spite of a negative GAS test.
Learn more in the ReAct Toolbox
Lessons to be learned
So what, if anything, can we learn from these studies? Hard conclusions can of course not be drawn based on single studies, especially when data is in conflict with previous publications. But some direction can nevertheless be extracted:
- The ultimate value of a diagnostic test is determined by its usefulness in improving antibiotic use. If the diagnostic doesn’t change antibiotic use, it is simply a waste of money, regardless of how good it is.
- A correct diagnosis means nothing if treatment guidelines are not followed. This is apparent from both studies. Despite results and guidelines indicating that antibiotics should not be used, antibiotics were prescribed in both studies. So new diagnostics are not automatically the solution for inappropriate antibiotic use – both prescribers and patients need to trust the diagnostic test.
- Investments in new diagnostic capacities need to consider the whole chain of diagnostics, not only a new test, no matter how advanced or easy to use. This ties in to both the above points. Even a perfect test is useless if it doesn’t change behavior. Thus, when introducing a new diagnostic test, it is important to ensure that any barriers to implementation of the test, from technology to eventual patient use of antibiotics are addressed. Every country or region will have their own drivers for inappropriate use of antibiotics, thus there is no one fix for all situations.
New reliable diagnostics for infectious diseases are needed for different reasons in different settings. It is crucial to identify the reasons for inappropriate use and to tailor each intervention accordingly to increase the impact of the interventions.