Demographic characteristics, bacterial isolates
Of the 1,448 participants we enrolled, females were the majority i.e. 63.3% (913/1,448). Thirty three percent of the participants (474/1,448) were from Kampala, 35% (508/1,448) from Kayunga and 32% (466/1,448) from Mpigi. The age range the participants 1 to 81 years, and mean of 25 years. Around 56% (802/1,448) of the participants were in the 15–44 year age group (Table S1). Of the 730 stool and 718 urine samples processed, 985 enterobacteria were isolated, of which 94.4% (930/985) were E. coli and 5.6% (55/985) were K. pneumoniae. Per district, the 985 enterobacteria were distributed as follows: 58% (274/474) Kampala, 80% (406/508) Kayunga, and 65.5% (305/466) Mpigi. The characteristics of the participants whose samples grew E. coli and K. pneumoniae are shown in Table S4. None of the urine samples grew bacteria at ≥ 104 colony forming units (CFU) per milliliter implying that there was no infection-related growth. Overall, 37% (535/1,448) of the participants visited outpatient clinics for general conditions and bacteria grew in 68% (363/535) of these participants. Of the 731 participants who presented with infectious conditions, 67% (488/731) had bacterial growth in their samples. Of the 122 participants who visited HIV/AIDS clinics for routine checks, 72% (88/122) had growth (Tables S1 & S2). Furthermore, 1,093 participants reported to have taken antibiotics 3 months prior to the visit, of whom 69% (755/1,093) had growth. Of the 125 participants who reported to have been previously admitted to hospitals, 62% (78/125) had growth. Of the 130 participants who reported to have undergone medical procedures 3 months prior to the clinic visit, 55% (71/130) had growth (Tables S1 & S2).
Prevalence of AmpC β-lactamase producing isolates
Of the 985 bacterial isolates investigated, 21% (209/985) were cefoxitin resistant. However, 25 cefoxitin resistant isolates were not available at the time of analysis, leaving 184 isolates that were investigated, of which 70% (129/184) were AmpC β-lactamase producers while 30% (55/184) were non-producers (Fig. 1). Therefore, the prevalence of AmpC β-lactamase producers among cefoxitin resistant isolates was 70% (129/184), implying the overall prevalence of AmpC β-lactamase producing isolates among enterobacteria was 13.1% (129/985); 12.5% (116/930 E. coli and 23.6% (13/55) Klebsiella. Per district the prevalence of AmpC β-lactamase producing bacteria was 23.7% (65/274) Kampala, 12.1% (37/305) Mpigi, and 6.7% (27/406) Kayunga. Furthermore, given the association between AmpC β-lactamases and clavulanate resistance, 247 amoxicillin/clavulanate resistant isolates in this study (see below) comprising of 229 E. coli and 18 Klebsiella, were tested for cefoxitin resistance, majority of which i.e. 84.6% (209/247) were found to be cefoxitin resistant while only 15.4% (38/247) were cefoxitin susceptible. Of the 209 amoxicillin/clavulanate and cefoxitin resistant isolates, 61.7% (129/209) were AmpC producers (Fig. 1).
Prevalence of pAmpC β-lactamase genes
One hundred and eleven of the 129 cefoxitin resistant and AmpC β-lactamase producing isolates were tested for pAmpC β-lactamase gene carriage. Of these, 60% (67/111) carried pAmpC genes. Furthermore, 47 of the 55 AmpC β-lactamase non-producers (see above) were tested for pAmpC gene carriage as they were cefoxitin resistant (MIC ≥ 16 µg/ml). Of these, 38% (18/47) carried pAmpC genes. Therefore, 54% (85/158) of the cefoxitin resistant isolates in this study (111 AmpC β-lactamase producers plus 47 AmpC β-lactamase non-producers) carried pAmpC β-lactamase genes. Isolates with reduced susceptibility to third-generation cephalosporins are suspects for AmpC β-lactamase production; in this study, 33.8% (22/65) of such isolates were cefoxitin susceptible, of which 59% (13/22) carried pAmpC genes. Overall, a total of 180 isolates (158 cefoxitin resistant plus 22 cefoxitin susceptible with reduced susceptibility to third-generation cephalosporins), comprising 157 E. coli and 23 Klebsiella, were tested for pAmpC gene carriage (Fig. 1). Of these, 54% (cefoxitin resistant isolates, 85/158) were pAmpC positive while 59.1% (cefoxitin susceptible isolates with reduced susceptibility to third-generation cephalosporins, 13/22) were pAmpC positive, giving a total of 98 pAmpC positive isolates investigated.
The overall prevalence of pAmpC genes in enterobacteria was 10% (98/985); by district it was 16.4% (45/274) Kampala, 6.2% (25/406) Kayunga and 9.2% (28/305) Mpigi, hence, the urban district of Kampala had more pAmpC gene positive isolates, Table 2. Per species the prevalence of pAmpC genes among cefoxitin resistant and/or amoxicillin/clavulanate resistant isolates was 59% (93/157) in E. coli and 26.1% (5/23) in Klebsiella. pAmpC β-lactamase gene carriage correlated with AmpC β-lactamase production (χ2 = 11.7, P-value 0.0003). The pAmpC β-lactamase producing E. coli belonged to phylogenetic groups A (n = 23), B1 (n = 10), B2 (n = 35) and D (n = 25). Overall, 39.6% (44/111) of AmpC β-lactamase producing isolates did not carry pAmpC genes, of which eight were Klebsiella that do not carry chromosomal AmpC genes. The AmpC β-lactamase producing isolates of E. coli that were pAmpC negative were assumed to be hyper-producers of chromosomal AmpC β-lactamases. Relatedly, the AmpC β-lactamase producing isolates of Klebsiella that were pAmpC negative likely carried genes we did not screen for.
The pAmpC genes detected were blaCIT (n = 54), blaCMY−2 (n = 23), blaCMY−4 (n = 31), blaEBC (n = 51, mainly blaACT−1) and blaDHA (n = 20), Table 2. Forty four isolates carried ≥ 2 pAmpC genes and the most frequent combination was blaEBC plus blaCIT. Furthermore, in this study, 11 ESBL producing isolates carried blaCTX−M−15 while three carried blaCTX−M−28. Co-existence of pAmpC with betalactamase genes (blaCTX−M, blaSHV, blaTEM) occurred in 49% (48/98), 9.2% (9/98) and 34.7% (34/98) isolates, respectively. Figure 2 depicts the distribution of β-lactamase genotypes in the three districts.
Table 2
Prevalence of pAmpC β-lactamase genes among E. coli and Klebsiella across the three districts
pAmpC genea | Specimen type (N = 98) | District | Totalb |
Stool 55 (56%) | Urine 43 (44%) | Kampala 45/274 (16.4%) | Kayunga 25/406 (6.2%) | Mpigi 28/305 (9.2%) |
blaCIT | 30 | 24 | 25 (13/12) | 12 (2/10) | 17 (8/9) | 54 |
blaDHA | 10 | 10 | 5 | 8 | 7 | 20 |
blaEBC | 28 | 23 | 28 | 11 | 12 | 51 |
aBased on a multiplex PCR by Perez-Perez and Hanson, 2002; bNumbers are derived from specimen type |
Antibiotic resistance patterns
The antibiotic resistance profiles of isolates investigated are depicted below (Table 2). Generally, cefoxitin resistance varied across the three districts and the variation was statistically significant (p = 0.023), Table 2. Furthermore, 59% (58/98) of the pAmpC β-lactamase producing isolates were resistant to a β-lactam antibiotic and to two other classes of commonly used non-β-lactam antibiotics, implying they were multidrug resistant (MDR) [29]. Two of the isolates were resistant to co-trimoxazole, ciprofloxacin, gentamicin, nitrofurantoin and chloramphenicol, while 7 were co-resistant to these drugs excluding nitrofurantoin, and one was resistant to the same drugs except chloramphenicol. Resistance to ciprofloxacin, co-trimoxazole and gentamicin was noted in 9 isolates, co-trimoxazole and chloramphenicol in 22 and co-trimoxazole and gentamicin in 5. Resistance to individual antibiotics among the pAmpC β-lactamase gene positive isolates was as follows, co-trimoxazole 79% (77/98); chloramphenicol 34.5% (34/98); ciprofloxacin 28% (27/98); gentamicin 23.5% (23/98); nitrofurantoin 8% (8/98); piperacillin/tazobactam 30.6% (30/98). None of the isolates was resistant to carbapenems.
Table 2
Antibiotic resistance rates among E. coli and Klebsiella isolated from individuals attending outpatient clinics in Kampala, Kayunga and Mpigi districts (2007–2008)
Antibiotic | Resistant (%) | P-value |
Kampalad n/N | Kayungae n/N | Mpigif n/N | Total n/N |
Ampicillina | 224/259 (86.5) | 218/396 (55.1) | 174/275 (63.3) | 616/930 (66.2) | 0.420 |
Amoxicillin/ clavulanate | 184/274 (67.2) | 80/406 (19.7) | 88/304 (29.0) | 352c/984 (35.8) | < 0.001 |
Cefuroxime | 43/273 (15.8) | 77/405 (19.0) | 22/304 (7.2) | 142/982 (14.5) | < 0.001 |
Ceftriaxone | 16/274 (5.8) | 2/406 (0.5) | 10/303 (3.3) | 28/983 (2.9) | < 0.001 |
Cefotaxime | 14/273 (5.1) | 2/406 (0.5) | 6/302 (2.0) | 22/981 (2.2) | < 0.001 |
Ceftazidime | 15/273 (5.5) | 2/406 (0.5) | 8/304 (2.6) | 25/983 (2.5) | < 0.001 |
Cefepimeb | 93/178 (52.3) | 26/76 (34.2) | 29/91 (31.9) | 148/345 (42.9) | 0.001 |
Ciprofloxacin | 73/274 (26.6) | 15/405 (3.7) | 17/304 (5.6) | 105/983 (10.7) | < 0.001 |
Sulfamethoxazole/Trimethoprim | 234/272 (86.0) | 262/405 (64.7) | 189/304 (62.2) | 685/981 (69.8%) | < 0.001 |
Gentamicin | 68/273 (24.9) | 19/406 (4.7) | 18/303 (5.9) | 105/982 (10.7) | < 0.001 |
Nitrofurantoin | 23/270 (8.5) | 4/403 (1.0) | 8/304 (2.6) | 35/977 (3.6) | < 0.001 |
Chloramphenicol | 83/273 (30.4) | 54/405 (13.3) | 56/301 (18.6) | 193/979 (19.7) | < 0.001 |
Cefoxitinb | 106/181 (58.6) | 40/78 (51.3) | 63/88 (71.6) | 209/347 (60.2) | 0.023 |
Piperacillin/ Tazobactumb | 126/173 (72.8) | 64/79 (81.0) | 67/93 (72.0) | 257/345 (74.5) | 0.315 |
Meropenem | 0/273 (0.0) | 0/403 (0.0) | 0/300 (0.0) | 0/976 (0.0) | - |
aRefers to E. coli; bOnly Amoxicillin/clavulanate resistant isolates tested; cRefers to inhibition zone diameter of ≤ 17 mm; dUrban district with a wet tropical climate; eRural district with a wet & dry tropical climate; fRural district with a wet tropical climate. |
Factors associated with carriage of pAmpC β-lactamase producing bacteria
There was significant association between Health Centre level, district of residence, use of ciprofloxacin and sample type with carriage of pAmpC β-lactamase producing bacteria, Table 3. After adjusting for each of them, the district of residence remained an independent risk factor for carriage of pAmpC β-lactamase producing bacteria, Table 3. Overall, we found that residing in a rural district (Kayunga/Mpigi) was associated with low carriage of pAmpC β-lactamase producing bacteria (aOR 0.23 (95% CI:0.11, 0.47) and aOR 0.49 (95% CI:0.25, 0.99), respectively. Similarly, participants who were 45 years and above carried less pAmpC positive bacteria (aOR 0.17 (95% CI:0.05, 0.62). Ciprofloxacin use was an independent risk factor for carriage of pAmpC positive bacteria (aOR 2.61 (95% CI:1.28, 5.32), Table 3.
Table 3
Factors associated with carriage of pAmpC β-lactamase producing bacteria
Characteristic | pAmpC gene |
Not present, n (%) | Present, n (%) | p-value | cOR (95% CI) | aOR (95% CI) |
Age group |
0–14 | 260 (31.8) | 36 (36.7) | 0.131 | 1.0 | 1.0 |
15–44 | 440 (53.9) | 55 (56.1) | 0.90 (0.58, 1.41) | 0.59 (0.32, 1.07) |
45+ | 117 (14.3) | 7 (7.2) | 0.43 (0.19, 1.00)* | 0.17 (0.05, 0.62)* |
Gender |
Female | 544 (66.5) | 59 (60.2) | 0.214 | 1.0 | 1.0 |
Male | 274 (33.5) | 39 (39.8) | 1.31 (0.85, 2.02) | 1.48 (085, 2.56) |
Health center level |
National Referral | 75 (9.1) | 18 (18.4) | 0.023 | 1.0 | 1.0 |
General Hospital | 220 (26.8) | 28 (28.6) | 0.53 (0.28, 1.01) | 1.04 (0.40, 2.70) |
Health Center IV | 134 (16.3) | 11 (11.2) | 0.34 (0.15, 0.76)* | 1.53 (0.47, 5.00) |
Health Center III | 392 (47.8) | 41 (41.8) | 0.44 (0.24, 0.80)* | 0.74 (0.28, 1.98) |
District |
Kampalaa | 198 (24.1) | 44 (44.9) | < 0.001 | 1.0 | 1.0 |
Kayungab | 363 (44.2) | 25 (25.5) | 0.31 (0.18, 0.52) | 0.23 (0.11, 0.47)* |
Mpigic | 260 (31.7) | 29 (29.6) | 0.50 (0.30, 0.83)* | 0.49 (0.25, 0.99)* |
Reason for visit |
ISS | 76 (9.8) | 5 (5.3) | 0.350 | 1.0 | 1.0 |
Infection | 398 (51.2) | 50 (52.6) | 1.91 (0.74, 4.94) | 2.60 (0.73, 9.32) |
General | 303 (39.0) | 40 (42.1) | 2.01 (0.77, 5.26) | 3.20 (0.86, 11.86) |
History of admission |
No | 748 (92.4) | 86 (88.7) | 0.207 | 1.0 | 1.0 |
Yes | 62 (7.6) | 11 (11.3) | 1.54 (0.78, 3.04) | 0.55 (0.21, 1.49) |
History of medical procedures |
Contact | 4 (7.3) | 1 (14.2) | 0.178 | | |
Inoculation | 40 (72.7) | 3 (42.9 | | |
Surgery | 11 (20.0) | 3 (42.9 | | |
Antibiotic use (any) |
No | 194 (23.6) | 24 (24.5) | 0.850 | | |
Yes | 627 (76.4) | 74 (75.5) | | |
Use of penicillin |
No | 427 (67.8) | 53 (73.6) | 0.313 | | |
Yes | 203 (32.2) | 19 (26.4) | | |
Use of Ciprofloxacin |
No | 566 (89.8) | 59 (80.8) | 0.020 | 1.0 | 1.0 |
Yes | 64 (10.2) | 14 (19.2) | 2.10 (1.11, 3.97) | 2.61 (1.28, 5.32) |
Use of co-trimoxazole |
No | 239 (34.3) | 34 (40.0) | 0.297 | | |
Yes | 458 (65.7) | 51 (60.0) | | |
cOR, crude odds ratio; aOR, adjusted odds ratio; ISS, immune suppression syndrome (HIV/AIDS); *Statistically significant association. |