Data from 2,571 patients analyzed, 1,610 patients from the NTH and 961 patients from the TH (Table 2). Antibiotics were commonly prescribed at both hospitals, although a significantly higher percentage of patients were prescribed antibiotics at the NTH (89% at the TH, 92% at the NTH, P < 0.05) (Table 2). Among each individual diagnosis group, there were no differences in percentage of antibiotic prescribing between the hospitals except for meningitis, were 90% of the patients were prescribed antibiotics at the NTH and 70% at the TH (P < 0.05). The number of antibiotic prescriptions per patient was higher at the TH compared to the NTH, among all patients the average number of antibiotic prescriptions (i.e. prescription of one specified antibiotic with stated dose, frequency and duration in days) per patient were 22 at the TH and 8 at the NTH (Table 2). Also, duration of hospital stay was higher among the patients at the TH compared to the NTH (mean 4.4 days at the NTH and 10.1 days at the TH, P < 0.05).
Table 2. Clinical characteristics and antibiotic prescribing among patients with severe infections at two private sector hospitals.
Patients
|
Number of patients, n (%)
|
Patients prescribed AB,
n (%)
|
Difference among patients prescribed AB,
OR (CI) p
P-value
|
Prescriptions
n (n of prescriptions per patients prescribed AB)
|
|
NTH
|
TH
|
NTH
|
TH
|
|
NTH
|
TH
|
All included patients
|
2504 (100)
|
1262 (100)
|
2294 (92)
|
1122 (89)
|
1.36 (1.08, 1.72) <0.05
|
18751 (8)
|
24956 (22)
|
Female
|
894 (36)
|
301 (24)
|
809(90)
|
259(86)
|
1.54 (1.01, 2.32) 0.03
|
|
|
Male
|
1610 (64)
|
961 (76)
|
1485(88)
|
863(90)
|
1.35 (1.01,1.80) 0.03
|
|
|
Cellulitis
|
388 (15)
|
402 (32)
|
354 (91)
|
362 (90)
|
1.15 (0.69, 1.91) 0.57
|
3505 (10)
|
8608 (24)
|
Endocarditis
|
7 (0)
|
2 (0)
|
6 (86)
|
1 (50)
|
6 (0.04, 547.49) 0.28
|
33 (6)
|
54 (54)
|
Epiglottitis
|
12 (0)
|
1 (0)
|
11 (92)
|
1 (100)
|
-
|
57 (5)
|
2 (2)
|
Meningitis
|
186 (7)
|
38 (3)
|
167 (90)
|
27 (71)
|
3.58 (1.37, 8.93) <0.05
|
1126 (7)
|
442 (16)
|
Peritonitis
|
431 (17)
|
252 (20)
|
402 (93)
|
233 (92)
|
1.13 (0.58, 2.14) 0.69
|
4909 (12)
|
6715 (29)
|
Pneumonia
|
761 (30)
|
410 (32)
|
692 (91)
|
366 (89)
|
1.21 (0.78, 1.83) 0.36
|
4686 (7)
|
6179 (17)
|
Pyelonephritis
|
71 (3)
|
3 (0)
|
68 (96)
|
2 (67)
|
11.33 (0.14, 262.11) 0.16
|
518 (8)
|
89 (45)
|
Septic arthritis
|
3 (0)
|
38 (3)
|
3 (100)
|
29 (76)
|
-
|
15 (5)
|
1159 (40)
|
Sepsis
|
645 (26)
|
116 (9)
|
591 (92)
|
101 (87)
|
1.6 (0.82, 3.0) 0.11
|
3902 (7)
|
1708 (17)
|
Notes: Statistically significant p-values are marked in bold font. Abbreviations: AB, antibiotic; CI, confidence interval; n, number; NTH, non-teaching hospital; OR, odds ratio; TH, teaching hospital
The overall antibiotic prescribing among all included patients increased from 2008 to 2017 at the NTH (P < 0.01) but did not significantly change at the TH (P = 0.07, Figure 1, Table 3). Antibiotics included in the access category comprised 61% of the total antibiotic prescribing at the TH and 40% at the NTH. Prescribing of access antibiotics increased at the NTH from 2008 to 2017 (P < 0.01, Figure 2, Table 3). Prescribing of antibiotics categorized as watch antibiotics comprised 29% of the total antibiotic prescribing at the TH and 40% at the NTH. Prescribing of watch antibiotics increased at both hospitals from 2008 to 2017 (P < 0.01 for both hospitals, Figure 2, Table 3). Reserve antibiotics comprised less than one percent of the antibiotic prescribing at both hospitals, however prescribing of reserve antibiotics increased from 2008 to 2017 at the TH (P < 0.01, Figure 2, Table 3). Prescribing of FDCs of antibiotics comprised 8% of the antibiotic prescribing at the TH and 18% at the NTH. Prescribing of FDCs of antibiotics increased at both hospitals from 2008 to 2017 (P < 0.01 for both hospitals, Figure 2, Table 3).
Table 3. Description of trends in antibiotic prescribing among patients with severe infections in Ujjain between 2008-2017.
|
NTH
|
TH
|
Antibiotic prescribing among specific diagnoses
|
All antibiotics
|
13.84 (< 0.01)
|
1.82 (0.07)
|
Cellulitis
|
5.72 (< 0.01)
|
6.52 (< 0.01)
|
Peritonitis
|
14.59 (< 0.01)
|
18.52 (< 0.01)
|
Pneumonia
|
4.87 (< 0.01)
|
7.30 (< 0.01)
|
Sepsis
|
2.18 (0.03)
|
-21.91 (< 0.01)
|
Antibiotic prescribing among all included patients
|
Access antibiotics
|
11.52 (< 0.01)
|
1.78 (< 0.07)
|
Watch antibiotics
|
9.63 (< 0.01)
|
6.49 (< 0.01)
|
Reserve antibiotics
|
-0.76 (0.45)
|
2.54 (< 0.01)
|
FDCs of antibiotics
|
14.28 (< 0.01)
|
7.31 (< 0.01)
|
Antibiotic prescribing among sepsis patients
|
Access antibiotics
|
1.49 (0.14)
|
-16.89 (< 0.01)
|
Watch antibiotics
|
3.02 (< 0.01)
|
-11.38 (< 0.01)
|
Reserve antibiotics
|
-9.32 (< 0.01)
|
Too few prescriptions
|
FDCs of antibiotics
|
3.78 (< 0.01)
|
-9.93 (< 0.01)
|
Notes: All values are presented with a value for the slope: t (obtained from linear regression analysis), followed by P-value in parenthesis. A positive t shows a positive trend and a negative t shows a negative trend. Statistically significant p-values are marked in bold font. Abbreviations: FDC, fixed dose combination; NTH, non-teaching hospital; TH, teaching hospital.
The most common diagnosis groups were cellulitis, peritonitis, pneumonia and sepsis that comprised 88% of the included patients at the NTH and 93% of the included patients at the TH. At both hospitals, there were few patients in the diagnosis groups: endocarditis, epiglottitis, meningitis pyelonephritis and septic arthritis (Table 2). The antibiotic groups comprising at least 75% of antibiotic prescribing within each of the four most common diagnoses; cellulitis, peritonitis, pneumonia or sepsis, as presented in Table 4. Total antibiotic prescribing among patients with cellulitis, peritonitis and pneumonia increased between 2008-2017 at both the NTH and the TH (P < 0.01 respectively, for both hospitals, Figure 1, Table 3). Total antibiotic prescribing among patients with sepsis increased from 2008 to 2017 at the NTH (P = 0.03), while it decreased at the TH (P < 0.01, Figure 1, Table 3). Prescribing of access, watch, reserve and FDCs of antibiotics among patients with sepsis is presented in Figure 3 and Table 3, since sepsis was the only of these diagnoses where the consumption decreased at one of the hospitals.
Table 4. Antibiotics prescribed among patients with severe infections at two private sector hospitals from 2007-2018.
|
Antibiotic prescribing in DDDs per 1000 patient days
|
%
|
ta
|
P-value
|
Year
|
2008
|
2009
|
2010
|
2011
|
2012
|
2013
|
2014
|
2015
|
2016
|
2017
|
|
|
|
Cellulitis NTH
|
J01CR
|
4.12
|
4.44
|
0.73
|
4.78
|
3.17
|
4.90
|
5.77
|
4.33
|
3.24
|
7.11
|
16
|
2.32
|
0.02
|
J01D
|
13.11
|
9.01
|
1.98
|
11.52
|
5.09
|
4.28
|
12.50
|
7.54
|
14.18
|
3.06
|
32
|
6.33
|
<0.01
|
J01R
|
6.53
|
5.60
|
3.12
|
2.60
|
4.70
|
3.46
|
3.01
|
6.78
|
6.12
|
1.84
|
17
|
6.97
|
<0.01
|
J01X
|
9.25
|
15.76
|
3.68
|
3.34
|
1.30
|
1.05
|
1.62
|
1.08
|
1.42
|
1.48
|
15
|
3.31
|
<0.01
|
Cellulitis TH
|
J01CR
|
10.84
|
7.45
|
4.24
|
7.14
|
9.47
|
8.99
|
10.07
|
8.03
|
16.29
|
13.33
|
19
|
4.91
|
<0.01
|
J01D
|
4.48
|
6.01
|
9.32
|
6.22
|
4.76
|
4.66
|
6.87
|
6.53
|
9.98
|
2.10
|
12
|
5.58
|
<0.01
|
J01G
|
9.33
|
13.53
|
9.11
|
9.26
|
14.55
|
10.47
|
10.01
|
6.37
|
15.64
|
5.45
|
21
|
0.44
|
0.66
|
J01M
|
10.20
|
8.70
|
3.35
|
5.66
|
9.00
|
4.51
|
4.30
|
3.31
|
9.11
|
6.22
|
13
|
2.66
|
<0.01
|
J01X
|
6.53
|
3.70
|
7.38
|
6.50
|
6.86
|
10.22
|
7.89
|
3.84
|
9.17
|
3.91
|
13
|
2.27
|
0.02
|
Peritonitis NTH
|
J01CR
|
3.71
|
4.26
|
3.85
|
10.23
|
3.06
|
5.70
|
4.45
|
5.89
|
5.67
|
4.40
|
16
|
4.98
|
<0.01
|
J01D
|
26.28
|
6.43
|
4.54
|
4.34
|
10.36
|
7.35
|
1.41
|
7.78
|
8.04
|
5.06
|
25
|
1.34
|
0.18
|
J01R
|
8.58
|
6.14
|
5.29
|
5.55
|
5.01
|
1.14
|
3.09
|
5.02
|
6.19
|
1.40
|
15
|
8.67
|
<0.01
|
J01X
|
16.03
|
11.72
|
11.64
|
9.09
|
4.57
|
2.49
|
1.58
|
1.81
|
3.54
|
1.97
|
20
|
10.95
|
<0.01
|
Peritonitis TH
|
J01CR
|
5.84
|
4.39
|
6.68
|
5.55
|
10.92
|
9.13
|
16.54
|
13.44
|
13.36
|
5.35
|
15
|
10.16
|
<0.01
|
J01D
|
4.00
|
4.95
|
9.63
|
9.44
|
6.87
|
2.76
|
6.04
|
7.64
|
15.87
|
7.73
|
12
|
17.55
|
<0.01
|
J01G
|
10.08
|
4.65
|
11.08
|
9.94
|
11.84
|
6.49
|
11.43
|
11.87
|
8.64
|
4.08
|
15
|
8.13
|
<0.01
|
J01M
|
16.05
|
10.72
|
3.27
|
6.40
|
11.99
|
5.83
|
11.60
|
12.86
|
15.42
|
3.56
|
16
|
2.92
|
<0.01
|
J01X
|
15.75
|
9.55
|
16.85
|
17.73
|
15.44
|
13.92
|
22.64
|
17.30
|
23.05
|
5.24
|
26
|
8.69
|
<0.01
|
Pneumonia NTH
|
J01CR
|
11.54
|
5.01
|
4.05
|
4.27
|
4.95
|
4.61
|
5.96
|
5.79
|
8.79
|
8.17
|
29
|
-0.91
|
0.36
|
J01D
|
7.75
|
6.92
|
4.22
|
8.53
|
3.24
|
2.95
|
1.37
|
4.44
|
4.23
|
4.18
|
22
|
-1.38
|
0.17
|
J01M
|
1.00
|
0.49
|
0.47
|
0.50
|
2.41
|
0.83
|
3.98
|
7.34
|
3.47
|
1.18
|
10
|
2.54
|
0.01
|
J01R
|
5.74
|
4.61
|
3.51
|
2.55
|
4.21
|
4.76
|
5.27
|
4.61
|
3.72
|
4.82
|
20
|
5.65
|
<0.01
|
Pneumonia TH
|
J01A
|
16.44
|
11.85
|
14.23
|
8.45
|
13.76
|
10.18
|
7.21
|
2.91
|
2.25
|
18.79
|
25
|
1.81
|
0.07
|
J01CR
|
8.45
|
12.64
|
10.66
|
1.91
|
8.26
|
15.71
|
20.02
|
19.21
|
15.38
|
13.27
|
30
|
6.65
|
<0.01
|
J01D
|
3.85
|
6.01
|
5.04
|
10.91
|
7.24
|
4.65
|
3.94
|
3.81
|
15.31
|
20.74
|
19
|
12.87
|
<0.01
|
J01M
|
15.87
|
9.36
|
8.72
|
5.72
|
3.88
|
2.04
|
1.61
|
1.77
|
4.65
|
7.44
|
14
|
1.89
|
0.06
|
Sepsis NTH
|
J01CR
|
5.05
|
3.90
|
2.13
|
1.97
|
2.23
|
2.55
|
3.01
|
6.90
|
3.47
|
6.66
|
21
|
1.10
|
0.27
|
J01D
|
10.74
|
7.47
|
1.19
|
3.27
|
4.63
|
3.86
|
4.77
|
6.37
|
4.29
|
1.70
|
27
|
-4.67
|
<0.01
|
J01R
|
7.55
|
3.66
|
1.71
|
1.33
|
2.09
|
3.36
|
1.84
|
3.07
|
2.63
|
2.76
|
17
|
3.17
|
<0.01
|
J01X
|
12.22
|
6.54
|
4.51
|
0.99
|
0.79
|
0.00
|
1.28
|
1.02
|
0.72
|
0.00
|
16
|
-6.07
|
<0.01
|
Sepsis TH
|
J01A
|
14.61
|
7.67
|
9.59
|
1.83
|
4.31
|
2.95
|
4.26
|
0.00
|
0.00
|
3.13
|
11
|
-5.50
|
<0.01
|
J01CR
|
1.84
|
11.18
|
27.87
|
2.76
|
5.09
|
1.83
|
8.04
|
7.09
|
7.79
|
7.09
|
19
|
-5.07
|
<0.01
|
J01D
|
1.37
|
5.48
|
8.48
|
3.03
|
5.87
|
3.79
|
12.27
|
5.57
|
7.21
|
4.89
|
13
|
-5.20
|
<0.01
|
J01M
|
11.51
|
1.75
|
14.45
|
11.32
|
3.13
|
0.11
|
4.75
|
0.91
|
2.81
|
3.62
|
13
|
-8.49
|
<0.01
|
J01X
|
8.95
|
12.93
|
15.34
|
7.08
|
9.18
|
6.36
|
12.60
|
0.91
|
6.00
|
4.07
|
19
|
-15.48
|
<0.01
|
Notes: Antibiotics presented comprise >75% of the total antibiotic prescribing within each diagnosis group. Numbers are presented in total antibiotic prescribing of antibiotic groups for each year, measured in DDD/1000 patient days and percentage of the total prescribing of antibiotic groups for each diagnosis and hospitals during the study period. a t obtained by linear regression. Statistically significant p-values are marked in bold font. Abbreviations: DDD, defined daily doses; NTH, non-teaching hospital; TH, teaching hospital.
Cellulitis
The WHO’s recommendations for empirical antibiotic treatment for cellulitis are combination of amoxicillin with clavulanic acid (J01CR) or cloxacillin (J01CF) as first choice and cefalexin (J01D) as second choice treatment [18]. At the NTH, the two most commonly prescribed antibiotics were of second choice treatment; J01D and J01R (FDCs of antibiotics). J01D antibiotics comprised 32% of antibiotic prescribing and J01R comprised 17% of the prescribing among patients with cellulitis. Prescribing of both J01D and J01R increased during the study period, and so did prescribing of recommended treatment J01CR and J01D (P < 0.05 for all, Table 4). At the TH the most commonly prescribed antibiotics among patients with cellulitis were from the J01CR (19%) and J01G group (21%), prescribing of J01CR increased from 2008 to 2017 (p < 0.01, Table 4). At the NTH, prescribing of access, watch and FDCs of antibiotics increased from 2008 to 2017 while at the TH, prescribing of access, watch, reserve and FDCs of antibiotics increased (P < 0.01 for all categories, at both hospitals).
Peritonitis
The World Society of Emergency Surgery’s recommendations (2017) for empirical antibiotic treatment of peritonitis are separate for community acquired or health care-associated peritonitis [21]. The first- and second choice antibiotics for community acquired peritonitis are presented in Table 5. At the NTH, the two most commonly prescribed antibiotics were of first choice treatment for community acquired peritonitis; J01D antibiotics comprised 25% of antibiotic prescribing and J01X 20% of the prescribing among patients with peritonitis, prescribing of J01X increased during the study period (P < 0.01, Table 4). At the TH, the two most commonly prescribed antibiotics were of first choice treatment for community acquired peritonitis; J01M (16%) and J01X group (26%), prescribing of both J01M and J01X increased from 2008 to 2017 (P < 0.01 for both antibiotic groups, Table 4). Further, prescribing of J01CR (first choice treatment for community acquired peritonitis), also increased during the study period at both hospitals. At both hospitals, prescribing of access, watch, reserve and FDCs of antibiotics increased from 2008 to 2017 (P < 0.01 for all categories mentioned at both hospitals).
Table 5. The World Society of Emergency Surgery’s recommendations for empirical antibiotic treatment of peritonitis.
|
First choice antibiotics
|
Second choice antibiotics
|
Community acquired peritonitis
|
Amoxicillin with clavulanic acid,
ceftriazone with metronidazole,
cefotaxime with metronidazole,
ciprofloxacin with metronidazole,
moxifloxacin.
For patients with risk of infection with extended spectrum beta lactamase (ESBL)-producing enterobacteriacea: ertapenem or tigecycline.
|
Meropenem,
doripenem,
imipenem with cilastatin,
ceftolozane with tazobactam and metronidazole, ceftazidime with avibactam and metronidazole plus additional vancomycin and teicoplanin. In patients at risk for infection with vancomycin-resistant enterococci: linezolid or daptomycin.
|
Health care associated peritonitis
|
Piperacillin with tazobactam,
meropenem with ampicillin,
doripenem with ampicillin,
imipenem with cilastatin,
tigecycline or amikacin.
|
Piperacillin with tazobactam,
cefepime with metronidazole.
In patients at risk for infection with ESBL-producing enterobacteriacea: meropenem, doripenem or imipenem with cilastatin.
|
Abbreviations: ESBL, extended spectrum beta lactamase.
Pneumonia
The WHO’s recommendations for empirical antibiotic treatment for pneumonia are presented in Table 6 [18]. At the NTH, the two most commonly prescribed antibiotics were of first choice treatment for community acquired pneumonia (J01D) and of second choice treatment for community/first choice treatment of health care acquired pneumonia (J01CR, Table 4). J01CR antibiotics comprised 29% and J01D 22% of the prescribing among patients with pneumonia, the prescribing did not change from 2008 to 2017. At the TH, the most commonly prescribed antibiotics among patients with pneumonia were of second choice treatment for community acquired pneumonia (J01A, 25%) and of second choice treatment for community/first choice treatment of health care acquired pneumonia (J01CR, 30%), prescribing J01CR increased from 2008 to 2017 (P < 0.01, Table 4). At the NTH, prescribing of access, watch and FDCs of antibiotics increased from 2008 to 2017. At the TH, prescribing of access and watch antibiotics increased from 2008 to 2017, while the prescribing of FDCs of antibiotics decreased (P < 0.01 for all categories at both hospitals).
Table 6. World Health Organization’s recommendations for empirical antibiotic treatment of pneumonia.
|
First choice antibiotics
|
Second choice antibiotics
|
Community acquired peritonitis
|
Ampicillin, benzylpenicillin, cefotaxime, ceftriaxone, clarithromycin, gentamicin
|
Amoxicillin with clavulanic acid, doxycycline
|
Health care associated peritonitis
|
Amoxicillin, amoxicillin with clavulanic acid, ampicillin, benzylpenicillin, cefotaxime, ceftriaxone, phenoxymethylpenicillin, piperacillin-tazobactam
|
|
Sepsis
The “Surviving sepsis campaign’s” international guidelines for empirical antibiotic treatment for sepsis and septic shock (2016) recommends: carbapenems (meropenem, imipenem/cilastatin or doripenem) or extended-range penicillin/β-lactamase inhibitor combination (piperacillin/tazobactam or ticarcillin/clavulanate) and third, or higher-generation cephalosporins can also be used [22]. At both hospitals, piperacillin with tazobactam from the J01CR group, were commonly prescribed, which is adherent to guidelines [22]. At the NTH, J01CR antibiotics comprised 21% of antibiotic prescribing and J01D 27% of the prescribing among patients with sepsis, prescribing of J01D decreased from 2008 to 2017 (P < 0.01, Table 4). At the TH, the most commonly prescribed antibiotics among patients with sepsis were from the J01CR (19%) and J01X group (19%), prescribing of both J01CR and J01X decreased from 2008 to 2017 (P < 0.01 for both antibiotic groups, Table 4). At the NTH, prescribing of watch and FDCs of antibiotics increased from 2008 to 2017 while prescribing of reserve antibiotics decreased. At the TH, prescribing in DDDs per 1000 patient days, of access, watch and FDCs of antibiotics decreased from 2008 to 2017 (Figure 3, Table 3).