Antimicrobial resistance and clinical effectiveness of co-trimoxazole versus amoxycillin for pneumonia among children in Pakistan: randomised controlled trial. Pakistan Co-trimoxazole Study Group.

BACKGROUND: Co-trimoxazole is widely used in treatment of paediatric pneumonia in developing countries, but drug resistance may decrease its effectiveness. We studied the effectiveness of co-trimoxazole compared with that of amoxycillin in pneumonia therapy, and assessed the clinical impact of co-trimoxazole resistance. METHODS: We recruited 595 children, aged 2-59 months, with non-severe or severe pneumonia (WHO criteria) diagnosed in the outpatient wards of two urban Pakistan hospitals. Patients were randomly assigned on a 2:1 basis co-trimoxazole (n=398) or amoxycillin (n=197) in standard WHO doses and dosing schedules, and were monitored in study wards. The primary outcome was inpatient therapy failure (clinical criteria) or clinical evidence of pneumonia at outpatient follow-up examination. FINDINGS: There were 92 (23%) therapy failures in the co-trimoxazole group and 30 (15%) in the amoxycillin group (p=0.03)-26 (13%) versus 12 (12%) among children with non-severe pneumonia (p=0.856) and 66 (33%) versus 18 (18%) among those with severe pneumonia (p=0.009). For patients with severe pneumonia, age under 1 year (p=0.056) and positive chest radiographs (p=0.005) also predicted therapy failure. There was no significant association between antimicrobial minimum inhibitory concentration and outcome among bacteraemic children treated with co-trimoxazole. INTERPRETATION: Co-trimoxazole provided effective therapy in non-severe pneumonia. For severe, life-threatening pneumonia, however, co-trimoxazole is less likely than amoxycillin to be effective.

Use of nasopharyngeal isolates of Streptococcus pneumoniae and Haemophilus influenzae from children in Pakistan for surveillance for antimicrobial resistance.

Antimicrobial resistance of Streptococcus pneumoniae and Haemophilus influenzae presents a challenge to clinical case management, particularly in programs for acute respiratory tract infection (ARI), including pneumonia, in developing countries. To determine whether nasopharyngeal isolates of S. pneumoniae and H. influenzae from a clinically defined group of children could be used to predict the prevalence of antimicrobial resistance of strains that cause disease, 601 urban children with ARI, 133 healthy urban children and 285 rural children were evaluated in Pakistan. Of the urban children with ARI, 216 (35.9%) were bacteremic, predominantly with S. pneumoniae (108 children) and H. influenzae (100 children). Overall 631 (61.9%) children carried S. pneumoniae and 381 (37.4%) carried H. influenzae. The proportions of nasopharyngeal isolates of both organisms from urban children with ARI resistant to penicillin or ampicillin, trimethoprim/sulfamethoxazole, chloramphenicol and erythromycin were similar to the proportions of resistant blood isolates. Nasopharyngeal isolates from rural children had lower rates of resistance to some antimicrobial agents. These findings suggest that nasopharyngeal isolates of S. pneumoniae and H. influenzae from children with ARI can be used to conduct surveillance for antimicrobial resistance in a defined geographic area. Such surveillance would aid programs in developing countries in making a rational choice of antimicrobial agents for use in clinical management of bacterial diseases, including pneumonia.

Antimicrobial resistance of pneumococci in children with acute lower respiratory tract infection in Pakistan.

87 strains of Streptococcus pneumoniae isolated during three winter seasons (1986-89) from the blood of children with acute lower respiratory tract infection (ALRI) in Pakistan were serotyped and tested for susceptibility to a range of antimicrobial agents. 97% of isolates were resistant to at least one antimicrobial drug. 62% had decreased susceptibility to co-trimoxazole (trimethoprim/sulphamethoxazole) (31% were fully resistant) and 39% were resistant to chloramphenicol. All isolates were susceptible to erythromycin, cefaclor, cephalothin, ceftriaxone, cefuroxime, rifampicin, vancomycin, and clindamycin. 29% of isolates were neither vaccine types nor vaccine-related types. Serotype distribution and antimicrobial susceptibility varied significantly during the three winter seasons. No single serotype was found in all three winters. The findings highlight the need for surveillance of antimicrobial resistance and serotype distribution of S pneumoniae in developing countries as a guide both to the choice of agent for treatment of pneumococcal infections, especially ALRI, and to the formulation of new pneumococcal conjugate vaccines for use in young children.

Diagnoses of acute lower respiratory tract infections in children in Rawalpindi and Islamabad, Pakistan.

A hospital-based inpatient and outpatient study of 1,492 cases of acute lower respiratory tract infection (ALRI) was conducted from November 1986 to March 1988 in two hospitals in Rawalpindi and Islamabad, Pakistan. Specimens of nasopharyngeal aspirate were processed for viral studies in all cases; blood cultures were performed in 1,331 cases; and urine was obtained for detection of bacterial antigen in 378 cases, but 227 of these samples had bacterial contamination and were discarded. Respiratory syncytial virus was identified in 33% of cases, and Haemophilus influenzae and Streptococcus pneumoniae were identified in 9.6% and 9.9% of cases, respectively. Nonencapsulated H. influenzae accounted for 32% of the Haemophilus isolates, and type b was the only encapsulated H. influenzae strain identified. Of the S. pneumoniae serotypes isolated, 31% are not included in the currently available polyvalent pneumococcal vaccine. No clinical characteristic was demonstrated to be a reliable indicator for bacterial ALRI.

Clonal analysis of Hemophilus influenzae isolated from children from Pakistan with lower respiratory tract infections.

The clonal diversity of 105 Hemophilus isolates from the blood of children with lower respiratory tract infection in Pakistan was analyzed. Ten isolates were identified as H. parainfluenzae and 95 as H. influenzae. Of the H. influenzae isolates, 61 (64%) were serotype b and 34 (36%) were nontypable; 95% of the type b isolates were members of a single clonal group (as defined by multilocus enzyme electrophoresis, SDS-PAGE outer membrane protein profile, and biotype). This clone is rarely observed among type b strains recovered from patients with invasive type b disease in the USA or Europe. The nontypable isolates in Islamabad also were clonally restricted: 9 clonal groups were found among 34 isolates, with just 5 clonal groups accounting for most (82%) of the strains. Children infected with type b strains were hospitalized more often than those with nontypable H. influenzae disease (64% vs. 41%, P = .06), but no other clinical or demographic features distinguished children infected by type b and nontypable strains.