In case of failureto diagnose inspite of rational approach, the following may be considered:

• Trial with chloroquine (considering epidemiology)

• Trials with antibiotics, choice based on the presumptive diagnosis. Trial with broad spectrum antibiotics

  should not be considered without relevant investigations to rule out serious infections such as UTI, Meningitis, Typhoid or Pneumonia.

• If the response to the first antibiotic is poor, another drug may be tried.

• If two drugs have failed, it is logical to reconsider the diagnosis rather than change the antibiotic.

• Do not try empirical treatment for tuberculosis except in life-threatening situations, wherein treatment must be completed for full conventional period, unless another cause for fever is found out during the trial period.

• Steroids should never be used for undiagnosed fever.

II. Empirical choice of antibiotic in fever:

• Do not prescribe an antibiotic without presumptive diagnosis.

• Routine investigations must be carried out to support the diagnosis.

• As clinical diagnosis of Bacterial infection in office practice is rarely possible within the first 2-3 days of fever, (except in case of Tonsillitis or otitis) prescribing antibiotic is not recommended during this period.

• If antibiotic is justified then, for most community infections, oral amoxycillin, or cotrimoxazole is sufficient (first line drugs).

• Injectable antibiotics are almost never needed in office practice.

• Newer antibiotics are not recommended for routine community acquired infections.

III. Change of specific antibiotic in fever:

• In confirmed diagnosis, change of therapy is rarely necessary.

• Improvement is anticipated over varying period of time in different drug sensitive, uncomplicated infections.

E.g. in malaria fever is under control in 2 days In respiratory infections 3-4 days In tuberculosis fever may continue for 1-2 weeks.

• In case of therapeutic failure check drug dosage, compliance, presence of any complications and the correctness of the diagnosis.

• In case the diagnosis is reasonably confirmed again, appropriate change of antibiotic is necessary based on the knowledge of prevalence of drug resistance pattern, unless specific bacteriological diagnosis and drug sensitivity pattern is available for choice of an antibiotic.

• Therapeutic failure even after the second trial of antibiotic demands reconsideration of diagnosis.

IV. Clinical approach to partially treated fever:

In such conditions the clinical picture might have been modified by prior therapy. If etiologic evidence is based on reasonable evidence, dosage & compliance of drug is checked. In case of suspected drug resistance, change of therapy is justified. If the disease has never been diagnosed and the therapy is empirical, failure of response may be due to wrong diagnosis. It is best to continue empirical therapy while investigations are repeated to arrive at the right diagnosis.

V. Interpretation of laboratory tests in partially treated fever:

• Lab tests need to be repeated in patients who continue to be febrile even after few days of therapy

• Some tests may be modified by therapy. E.g. WBC counts, peripheral smear for malarial parasites, urinalysis and bacterial culture.

• Persistence of high ESR inspite of treatment suggests uncontrolled active disease. So is persistent eosinopenia, hence change in therapy may be indicated.

• Change from neutrophilic response to lymphocytic response in peripheral smear indicates recovering bacterial infection, hence to continue the same antibiotic.

• Improving laboratory tests with no clinical response should alert the physician to the possibility of complications.

• Clinical improvement with persistent abnormal tests warrants close observation without change in the antibiotic.