Antimicrobial resistance has become a defining patient-safety problem in modern hospitals. The immediate cause is biological selection under antimicrobial exposure, but the clinical expression of resistance is shaped by prescribing volume, spectrum, duration, infection-control practices, movement of patients, and the local ecology of pathogens. In the intensive care unit (ICU), these forces converge. Patients are severely ill, invasive devices are common, organ function changes rapidly, and diagnostic uncertainty is often greatest at the moment treatment decisions are most urgent. The result is a setting in which timely antimicrobial therapy can be lifesaving, yet unnecessary or poorly reviewed treatment can create substantial harm [1].

Infection is frequent among critically ill patients and is associated with longer stay, organ dysfunction, and mortality. Large international ICU datasets have shown that infection and antimicrobial exposure are widespread, with resistant organisms contributing appreciably to adverse outcomes [2]. The first therapeutic decision therefore has two competing obligations. It must provide adequate early coverage for a patient who may deteriorate rapidly, and it must avoid exposing patients and the ICU microbial environment to avoidable broad-spectrum treatment. Inadequate empirical therapy has been associated with higher mortality among critically ill patients [3], while delay in effective treatment during septic shock has also been linked to poorer survival [4]. Rational antimicrobial use does not mean withholding treatment when infection is likely. It means selecting, dosing, monitoring, reviewing, narrowing, and stopping therapy with discipline.

Antimicrobial stewardship is the organised expression of rational antimicrobial use. It combines clinical judgment with microbiology, pharmacology, infection control, audit, and feedback. The core objective is to obtain the best possible clinical outcome while minimising toxicity, unnecessary cost, ecological pressure, and the emergence of resistance. Early guidance described stewardship in terms of optimal drug selection, dose, route, and duration [5]. Later implementation guidelines emphasised leadership commitment, prospective audit and feedback, formulary restriction where appropriate, facility-specific treatment guidance, dose optimisation, and systematic measurement of antimicrobial use [6].

The ICU needs a tailored stewardship approach because standard ward-based rules may not fully address altered pharmacokinetics, augmented renal clearance, renal replacement therapy, fluid shifts, vasopressor use, mechanical ventilation, or the consequences of missing early effective treatment. ICU stewardship therefore begins with prompt and adequate empirical therapy when clinically indicated, but it must continue through a structured reassessment once cultures, imaging, biomarkers, organ function, and the clinical trajectory become clearer [7,8]. Broad empirical coverage should be viewed as a temporary bridge to targeted therapy, not as a default regimen to be continued without review.

Drug-utilisation studies provide a practical description of how medicines are actually used in routine care. They can identify polypharmacy, excessive reliance on injectable therapy, dominance of particular antimicrobial classes, limited de-escalation, prolonged treatment, and avoidable adverse effects. Such observations are especially valuable in institutions where electronic prescribing data, defined daily dose metrics, or unit-specific antibiograms are not routinely available. Antibiotic consumption and resistance are positively related at patient and population levels [9], while stewardship interventions can improve guideline adherence, reduce antimicrobial exposure, lower costs, and decrease the incidence of resistant organisms and Clostridioides difficile infection [10-13].

Hospital records Mahavir Institute of Medical Sciences were reviewed to assess antimicrobial prescribing among 400 adults admitted to intensive care units. The analysis covered patient demographics, admission diagnoses, comorbidities, number and class of antimicrobials, route and duration of therapy, drug combinations, suspected adverse reactions, treatment modifications, de-escalation, clinical improvement, referral, discharge against medical advice, and mortality. As culture and susceptibility findings were not available, antimicrobial resistance was not measured directly. The study therefore focused on prescribing practices that may contribute to resistance, with the objectives of evaluating the rationality of antimicrobial use, examining associations between polypharmacy and adverse reactions, identifying gaps in current prescribing, and proposing practical stewardship measures to reduce unnecessary antimicrobial exposure. measures for the ICU.

MATERIALS AND METHODS

Study design and setting

This was a prospective observational and descriptive drug-utilisation study conducted in the intensive care units of Mahavir Institute of Medical Sciences Shivareddypet,  Vikarabad,  India. The hospital functions as a tertiary teaching centre and receives critically ill medical, surgical, trauma, toxicology, obstetric, and emergency referrals. The study focused on prescribing practice during routine ICU management and did not alter the treating team’s therapeutic decisions.

Study period and study population

The study was conducted over 12 months, from April 2025 to March 2026. The study population consisted of adult patients admitted to the ICUs who received one or more antimicrobial agents during their stay. A total of 400 consecutive eligible patients were included. Each patient was observed from enrolment until discharge, referral, discharge against medical advice, or death, according to the information available in the source record.

Eligibility criteria

Patients older than 18 years of either sex who were admitted to an ICU and prescribed at least one antimicrobial were eligible. Pregnant or lactating women, patients younger than 18 years, individuals who declined participation, cases with insufficient clinical or prescribing information, and patients with systemic or local malignancy were excluded. These criteria were retained from the original study protocol.

Sampling and sample size

Purposive consecutive sampling was used. The prespecified sample size was 400 cases. All consecutive patients meeting the eligibility criteria during the study period were recruited until the required sample was achieved. The available source document did not provide a formal prevalence-based sample size calculation, and no retrospective power calculation was undertaken because the study was primarily descriptive.

Data collection procedure

Data were collected personally from ICU case records and entered in a preapproved case-record form. The variables included age, sex, admission diagnosis, inpatient number, admission and exit dates, comorbid conditions, laboratory information recorded in the case sheet, antimicrobial name, drug class, dose, frequency, route, duration, use of generic or trade names, monotherapy or combination therapy, reason for change, suspected adverse drug reactions, drug interactions when documented, and clinical disposition. Written informed consent was obtained from the patient or an authorised relative after the purpose and procedures had been explained.

Operational definitions and study outcomes

Polypharmacy was operationally classified as prescription of five or more total drugs during ICU care, matching the categorisation used in the source results. Antimicrobial monotherapy indicated use of one antimicrobial agent, while multitherapy indicated concurrent use of two or more agents. De-escalation referred to documented narrowing of therapy or conversion to an oral regimen. Suspected adverse drug reactions were recorded as documented by the treating team; the available source data did not include formal causality, severity, or preventability assessment. The main descriptive outcomes were antimicrobial selection, route, duration, regimen modification, adverse reactions, and final clinical disposition.

Ethical considerations

Institutional Ethics Committee clearance and administrative permission were obtained before recruitment. Written informed consent was obtained from patients or relatives.

Statistical analysis

Data were entered in Microsoft Excel and analysed using IBM SPSS Statistics for Windows, version 28.0 (IBM Corp., Armonk, New York, USA). Categorical variables were summarised using frequencies and percentages. Pearson’s chi-square test was used to examine associations between admission category and sex, total number of drugs and suspected adverse drug reactions, sex and clinical outcome, and adverse drug reactions and clinical outcome. A two-sided p value below 0.05 was considered statistically significant. Because age was available only in categories, no mean age was estimated. The median duration of therapy and interquartile range were derived from the recorded day-wise frequency distribution.

Data verification and handling of source inconsistencies

The numerical tables in the attached results were checked arithmetically before manuscript preparation. The adverse-reaction-by-outcome table listed the total number of patients with adverse reactions as 66, although its component cells summed to 241; the reconstructed analysis therefore used 241 and yielded χ²=6.12 with p=0.106. The antimicrobial table listed piperacillin-tazobactam use as 32 men and 17 women, but gave a total of 43. The sex-specific counts sum to 49, and the reconstructed table reports 49 with an explanatory footnote. These corrections are transparent arithmetic reconciliations rather than new observations. No culture, susceptibility, defined daily dose, days-of-therapy, severity-score, or infection-source data were added because they were not present in the supplied files.

RESULTS

Demographic characteristics

All 400 enrolled patients were included in the descriptive analysis. The largest age group comprised patients older than 50 years, accounting for 152 cases (38.0%). Patients aged 21-30 years formed the next largest group, with 90 cases (22.5%). There were 229 men (57.3%) and 171 women (42.8%). The complete age and sex profile is shown in Table 1, and the age distribution is illustrated in Figure 1.

Table 1: Demographic profile of the study population (n=400)

Figure 1: Age distribution of adult ICU patients receiving antimicrobial therapy. The oldest category contained the largest proportion of participants.

Reasons for ICU admission and comorbid conditions

Cardiovascular disorders were the most frequent admission category, affecting 79 patients (19.8%), followed by trauma in 54 (13.5%), poisoning in 44 (11.0%), central nervous system disorders in 35 (8.8%), and both respiratory and abdominal or gastrointestinal disorders in 34 patients each (8.5%). Obstetric and gynaecological conditions accounted for 32 admissions and occurred exclusively among women. Trauma and snake bite were predominantly recorded among men. The distribution of admission categories differed significantly by sex (χ²=65.06, df=9, p<0.001), as presented in Table 2 and Figure 2.

No comorbid illness was documented in 276 patients (69.1%). Hypertension alone was present in 49 (12.3%), hypertension with diabetes in 29 (7.3%), and diabetes alone in 20 (5.0%). Overall, 124 patients (30.9% as reported in the source discussion) had at least one recorded comorbid condition. The comorbidity profile is included in Table 2.

Table 2: Clinical profile, admission category by sex, and comorbidity status

GIT, gastrointestinal; CNS, central nervous system; CVS, cardiovascular system. The chi-square test compares the full 10 x 2 admission-category-by-sex table. Comorbidity percentages reproduce the source table and may differ slightly from direct recalculation because of source rounding. No inferential test was applied to the comorbidity distribution.

Figure 2: Reasons for admission to the ICU stratified by sex. The significant overall association reflects the female-only obstetric and gynaecological category and the male predominance in trauma and snake bite admissions

Overall prescribing burden and antimicrobial strategy

Five or more total drugs were prescribed to 306 patients (76.5%), indicating a high burden of concurrent pharmacotherapy in the ICU. Ninety-four patients (23.5%) received fewer than five drugs. Antimicrobial monotherapy was used in 273 cases (68.3%), while 127 patients (31.8%) received two or more antimicrobial agents. These combinations were generally used to broaden empirical coverage or address suspected mixed aerobic and anaerobic infection. The core prescribing indicators are summarised in Table 3 and Figure 3.

Table 3: Drug-utilisation and antimicrobial prescribing indicators

 Descriptive categories are not mutually exclusive across sections. No inferential test was applied to these single-variable distributions. IV, intravenous.

Figure 3: Selected prescribing indicators. Near-universal intravenous use and the very low documented de-escalation rate identify major opportunities for structured antimicrobial review

Route, duration, and modification of antimicrobial therapy

The intravenous route was used exclusively in 397 patients (99.25%). Two patients received oral treatment alone, and one received both intravenous and oral therapy. The median recorded treatment duration was 3 days, with an interquartile range of 2-5 days. A total of 384 patients (96.0%) received antimicrobials for seven days or less, while 16 (4.0%) had treatment extending beyond seven days. The short recorded duration in many cases reflected early referral, discharge against medical advice, or termination of observation rather than necessarily planned completion of an antimicrobial course.

The original antimicrobial regimen was continued without change in 391 patients (97.8%). Six regimens (1.5%) were modified because of inadequate clinical response, and three (0.8%) were documented as de-escalation or oral switch. No change was attributed to allergy or a severe adverse reaction in the available records. The detailed duration distribution is shown in Table 4.

Table 4: Duration of antimicrobial treatment

 Duration represents observed antimicrobial exposure in the source record. The source table reports 86 one-day courses as 21.75%, although direct calculation gives 21.50%; the reported percentage has been retained transparently

Antimicrobial classes and individual agents

Cephalosporins dominated prescribing. Cefotaxime was recorded in 276 patients (69.0%), making it the most frequently used agent by a wide margin. Metronidazole was used in 116 (29.0%), piperacillin-tazobactam in 49 (12.3%) after arithmetic reconstruction of the sex-specific counts, ceftriaxone and gentamicin in 28 patients each (7.0%), amoxicillin-clavulanate and ciprofloxacin in 20 each (5.0%), and cefoperazone-sulbactam in seven (1.8%). Meropenem and amikacin were each used in three patients, while azithromycin was used in two. Levofloxacin and fluconazole were each recorded once. Table 5 presents the individual agents, and Figure 4 displays the ten most frequently prescribed drugs.

Table 5: Class and individual antimicrobial agents used

Percentages use 400 patients as the denominator and may sum to more than 100 because patients receiving combination therapy could appear under more than one agent. The source table reported piperacillin-tazobactam as 32 men and 17 women but gave a total of 43; the arithmetically correct total of 49 is shown.

Figure 4: Most frequently prescribed antimicrobial agents. Cefotaxime accounted for the greatest exposure, followed by metronidazole

Suspected adverse drug reactions

At least one suspected adverse drug reaction was documented in 241 patients (60.2%), while 159 (39.8%) had no recorded reaction. Nausea or vomiting was the most frequent reaction, occurring in 66 patients (16.5%). Headache was recorded in 51 (12.8%), diarrhoea in 48 (12.0%), and altered taste sensation in 46 (11.5%). Less frequent events included pruritus or rash, thrombophlebitis, urticaria, anorexia, and photosensitivity. The reaction profile is presented in Table 6 and Figure 5. The available records did not include a formal causality scale, severity grading, or preventability assessment, so these events should be interpreted as documented suspected reactions rather than confirmed antimicrobial-related events.

Table 6: Documented suspected adverse drug reactions

The categories sum to 400 because one principal reaction category was recorded for each patient in the source table. No causality, severity, or preventability assessment was available

Figure 5: Distribution of documented suspected adverse drug reactions among patients receiving antimicrobial therapy

Clinical disposition

Clinical improvement with subsequent discharge after stabilisation was documented in 192 patients (48.0%). Eighty-nine patients (22.3%) were discharged against medical advice, and 66 (16.5%) were referred to a higher centre because the required facilities were unavailable locally. Fifty-three patients died, giving an observed mortality of 13.3%. Since 155 patients were either referred or discharged against medical advice, the final efficacy of antimicrobial treatment could not be determined for a substantial proportion of the cohort. Table 7 and Figure 6 show the clinical disposition.

Table 7: Clinical disposition at the end of observation

Disposition is not equivalent to antimicrobial efficacy. Outcomes were influenced by underlying illness, referral, availability of facilities, and discharge against medical advice

Figure 6: Clinical disposition at the end of observation. Almost two-fifths of the cohort had incomplete local follow-up because of referral or discharge against medical advice

Association analyses

Suspected adverse reactions were more frequent among patients prescribed five or more total drugs. Among patients receiving fewer than five drugs, 42 of 94 (44.7%) had a documented reaction, compared with 199 of 306 (65.0%) among those receiving five or more drugs. The association was statistically significant (χ²=12.44, df=1, p<0.001). This finding supports medication review as an important safety component of rational ICU prescribing.

Clinical outcome did not differ significantly by sex (χ²=2.30, df=3, p=0.513). The association between adverse-reaction status and outcome was also not statistically significant (χ²=6.12, df=3, p=0.106). These analyses are summarised in Table 8.

Table 8: Inferential analyses of prescribing safety and clinical outcome

ADR, suspected adverse drug reaction; DAMA, discharged against medical advice. The ADR-present total was reconstructed as 241 from the component cells. Pearson chi-square tests were two-sided; p<0.05 was considered significant.

DISCUSSION

This prospective ICU drug-utilisation study identified a prescribing pattern characterised by extensive polypharmacy, near-universal intravenous antimicrobial use, dominance of broad-spectrum cephalosporins, and very limited documented de-escalation. Cefotaxime was used in more than two-thirds of patients, while metronidazole was the second most frequent agent. Although most patients received one antimicrobial, almost one-third received combinations. The majority of observed courses lasted seven days or less, but the short durations must be interpreted in the context of referral and discharge against medical advice. Suspected adverse reactions were common in the record and were significantly associated with the total number of drugs prescribed. These findings point to clear opportunities for rationalisation, but they do not establish the prevalence or microbiological pattern of antimicrobial resistance because culture and susceptibility data were not available.

Patients older than 50 years formed the largest age category. This is clinically plausible because ageing is accompanied by greater cardiovascular, respiratory, neurological, and metabolic vulnerability, as well as reduced physiological reserve. Cardiovascular disease was the most common reason for ICU admission, followed by trauma and poisoning. The admission profile shows that antimicrobial therapy was prescribed not only for established infection but also in clinical contexts where prophylaxis or concern for secondary infection may have influenced treatment. This distinction is important. Antimicrobial need in perforation, sepsis, or severe respiratory infection is different from routine prophylaxis after trauma, poisoning, stroke, or monitoring admissions. A rational-use programme should therefore require a clearly documented indication for every antimicrobial order.

The significant association between sex and admission category was mainly structural. Obstetric and gynaecological admissions occurred only among women, while trauma and snake bite were predominantly male. Sex itself was not associated with clinical outcome. The finding should therefore not be interpreted as a biological difference in antimicrobial response. It reflects the case mix of a tertiary centre serving medical, surgical, obstetric, and emergency populations.

More than three-quarters of patients received at least five drugs. Polypharmacy is expected in intensive care because organ support, analgesia, sedation, thromboprophylaxis, fluid and electrolyte correction, glycaemic control, and treatment of comorbid illness are often required simultaneously. Nevertheless, every additional medicine increases the possibility of drug-drug interactions, duplicated therapy, administration error, renal or hepatic accumulation, and difficulty attributing an adverse event to a specific agent. In the present dataset, suspected adverse reactions occurred in 65.0% of patients receiving five or more drugs compared with 44.7% among those receiving fewer medicines. The significant association does not prove causality, but it supports daily multidisciplinary medication review.

Antibiotic exposure itself is associated with gastrointestinal, dermatological, renal, neurological, haematological, and other adverse events. A hospital cohort reported that longer antibiotic exposure increased the risk of antibiotic-associated adverse drug events [14]. The present record showed nausea or vomiting, headache, diarrhoea, and altered taste as the most common reactions. Because the source files did not contain causality assessment, some events may have arisen from critical illness or concomitant drugs. Future pharmacovigilance should use a standard causality tool, severity classification, preventability assessment, and details of management and outcome.

Cephalosporin predominance was the most striking prescribing feature. Cefotaxime alone was recorded in 69.0% of patients. Such heavy reliance on one broad-spectrum class may be understandable when rapid empirical coverage is required and microbiology support is delayed, but it can also produce strong ecological pressure. Third-generation cephalosporin exposure has implications for selection of extended-spectrum beta-lactamase-producing organisms and Clostridioides difficile. Stewardship programmes have been associated with reductions in resistant Gram-negative bacteria, methicillin-resistant Staphylococcus aureus, and C. difficile infection [11,15]. The local objective should not be to prohibit cephalosporins, but to reserve them for documented indications, align empirical choices with a current ICU antibiogram, and narrow treatment when culture results permit.

Metronidazole use in 29.0% suggests frequent empirical anaerobic coverage, particularly in abdominal, surgical, trauma, and obstetric settings. Combination treatment can be rational when polymicrobial infection is likely, but routine addition without a defined source may duplicate coverage or extend treatment unnecessarily. Piperacillin-tazobactam, amoxicillin-clavulanate, cefoperazone-sulbactam, aminoglycosides, and fluoroquinolones were also used. Each class carries distinct dose, toxicity, and resistance considerations. Renal function, body weight, infection site, organ support, and local susceptibility should guide dosing rather than fixed regimens alone.

Almost every patient received intravenous therapy. The intravenous route is often appropriate at the start of ICU treatment because of shock, altered gastrointestinal absorption, mechanical ventilation, reduced consciousness, and the need for reliable tissue exposure. However, continuing intravenous treatment after haemodynamic and gastrointestinal recovery exposes patients to line-related complications, nursing workload, and avoidable cost. Evidence syntheses identify intravenous-to-oral conversion as a stewardship objective associated with clinical and resource benefits [10]. A daily switch assessment should therefore be incorporated once the patient is stable, an active oral agent is available, and absorption is reliable.

Most observed courses lasted no more than seven days, and the median recorded duration was three days. A short course can be an indicator of rational use when it represents an intentional evidence-based stop decision. In this dataset, however, many one- to five-day observations were interrupted by referral or discharge against medical advice. Duration should therefore be interpreted as observed exposure rather than a completed course. Biomarker-guided discontinuation has reduced antibiotic exposure in critically ill patients without compromising safety in large randomised trials [18,19]. Biomarkers cannot replace clinical assessment or microbiology, but when available they can support a structured decision to stop therapy.

Only 0.8% of patients had documented de-escalation or oral switch, and 97.8% continued the initial regimen. This is the clearest stewardship gap in the study. An empirical regimen that is reasonable during the first hour of sepsis may become unnecessarily broad after 48 to 72 hours. At that point, the team should review cultures, susceptibility, imaging, source control, clinical response, organ function, and the probability of a non-infectious diagnosis. Observational evidence has associated de-escalation with lower mortality in severe sepsis and septic shock [16], while a randomised trial found that formal de-escalation was feasible but did not automatically shorten total exposure, highlighting the need for careful implementation [17].

The source records did not include culture positivity, organisms isolated, susceptibility patterns, time to culture collection, or the proportion of cultures obtained before antimicrobial administration. This limitation prevents assessment of concordance between empirical and definitive treatment. It also prevents construction of an ICU antibiogram, which is essential for rational empirical selection. Future audits should record the indication, infection source, culture timing, empirical regimen, microbiological result, definitive regimen, and reason for continuation or discontinuation. De-escalation should be documented as a clinical decision even when cultures are negative and treatment is stopped.

Improvement was documented in 48.0%, and mortality was 13.3%. These figures should not be interpreted as the efficacy or failure rate of antimicrobial therapy. Outcome in the ICU depends on illness severity, timeliness of source control, organ support, comorbidity, poisoning or trauma severity, referral decisions, and availability of higher-level services. Moreover, 38.8% of patients were referred or discharged against medical advice, so final outcomes were not available locally. The lack of APACHE II, SOFA, infection severity, and microbiological adequacy data also prevents adjusted outcome analysis.

The clinical imperative remains to avoid both undertreatment and overtreatment. Sepsis guidelines support prompt antimicrobials when infection is likely or septic shock is present, followed by daily assessment for de-escalation and appropriate duration [20]. For hospital-acquired and ventilator-associated pneumonia, guideline-based selection and seven-day treatment for many patients are recommended, with adaptation to clinical response and local susceptibility patterns [21]. These principles reconcile the need for early effective therapy with the need to limit unnecessary exposure.

The findings support a practical ICU stewardship programme centred on the prescribing event and a mandatory reassessment. At initiation, the prescriber should document the suspected source, severity, reason for empirical coverage, cultures obtained, planned duration, and dose adjustment. The antimicrobial order should carry a review or stop date. At 48 to 72 hours, a structured antibiotic timeout should classify the regimen as stop, continue unchanged with justification, narrow, broaden, change for toxicity, or convert from intravenous to oral therapy. The review should be visible in the case record and discussed during rounds.

A multidisciplinary stewardship team should include an intensivist, microbiologist, clinical pharmacologist or infectious-disease physician where available, pharmacist, infection-control nurse, and hospital administrator. The team should prepare an ICU-specific antibiogram at least annually, audit high-use agents such as cefotaxime, monitor days of therapy per 1,000 patient-days, review prolonged or duplicate anaerobic coverage, and provide feedback to prescribers. Education should be linked to actual local prescribing data rather than delivered as isolated lectures. The proposed cycle is summarised in Table 9 and Figure 7.

Table 9. Proposed ICU antimicrobial stewardship actions based on the observed prescribing pattern

These are evidence-informed recommendations generated from the observed prescribing gaps. They were not tested as an intervention in the present observational study. DAMA, discharge against medical advice

Figure 7: Proposed antimicrobial stewardship pathway for the ICU. The cycle preserves prompt empirical treatment while creating explicit checkpoints for microbiology, dose optimisation, de-escalation, route conversion, and audit.

The study included a relatively large consecutive cohort of 400 adult ICU patients and captured routine prescribing across diverse medical, surgical, trauma, toxicology, respiratory, cardiovascular, and obstetric presentations. It described the complete medication burden, antimicrobial route, duration, individual agents, treatment modification, suspected adverse reactions, and disposition. The significant association between polypharmacy and documented adverse reactions provides a clinically relevant safety signal. The manuscript also transparently reconciles arithmetic inconsistencies in the source tables instead of silently reproducing them.

The study was conducted at a single centre and used purposive consecutive sampling. The source files did not include culture and susceptibility results, organism distribution, infection source confirmation, timing of antimicrobial administration, surgical prophylaxis duration, defined daily doses, days of therapy, cost, generic-prescribing indicators, or adherence to a local guideline. The study therefore cannot directly quantify antimicrobial resistance or determine the appropriateness of each individual prescription.

Severity scores, renal and hepatic function at the time of dosing, mechanical ventilation, invasive-device exposure, source-control procedures, ICU length of stay, and adjusted mortality predictors were not available. Suspected adverse reactions were not assessed with a formal causality or severity scale. Many patients were referred or discharged against medical advice, which limited outcome ascertainment. Finally, the study was observational and did not include a pre-post stewardship intervention, so the proposed promotion strategy requires prospective implementation and evaluation.

CONCLUSION

Antimicrobial prescribing in the ICUs of BRIMS Teaching Hospital was characterised by a high overall drug burden, extensive use of intravenous therapy, marked reliance on cefotaxime, and minimal documented de-escalation. Most patients received one antimicrobial and had an observed treatment duration of seven days or less, but interruption by referral or discharge against medical advice limited assessment of completed courses. Suspected adverse reactions were common and were significantly associated with prescription of five or more total drugs. The data support introduction of a structured ICU antimicrobial stewardship programme built around clear indications, cultures before treatment when feasible, local antibiogram-guided empirical choices, dose optimisation, a mandatory 48-72-hour review, early intravenous-to-oral conversion, documented stop dates, pharmacovigilance, and regular audit with feedback. Because resistance and microbiological adequacy were not measured in the available dataset, future evaluation should link antimicrobial use to culture results, days of therapy, resistance trends, clinical severity, and risk-adjusted outcomes.

REFERENCES

1. Holmes AH, Moore LSP, Sundsfjord A, Steinbakk M, Regmi S, Karkey A, et al. Understanding the mechanisms and drivers of antimicrobial resistance. Lancet. 2016;387(10014):176-187. doi:10.1016/S0140-6736(15)00473-0.
2. Vincent JL, Rello J, Marshall J, Silva E, Anzueto A, Martin CD, et al. International study of the prevalence and outcomes of infection in intensive care units. JAMA. 2009;302(21):2323-2329. doi:10.1001/jama.2009.1754.
3. Kollef MH, Sherman G, Ward S, Fraser VJ. Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest. 1999;115(2):462-474. doi:10.1378/chest.115.2.462.
4. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med. 2006;34(6):1589-1596. doi:10.1097/01.CCM.0000217961.75225.E9.
5. Dellit TH, Owens RC, McGowan JE Jr, Gerding DN, Weinstein RA, Burke JP, et al. Infectious Diseases Society of America and Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis. 2007;44(2):159-177. doi:10.1086/510393.
6. Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, et al. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016;62(10):e51-e77. doi:10.1093/cid/ciw118.
7. Arnold HM, Micek ST, Skrupky LP, Kollef MH. Antibiotic stewardship in the intensive care unit. Semin Respir Crit Care Med. 2011;32(2):215-227. doi:10.1055/s-0031-1275534.
8. Luyt CE, Brechot N, Trouillet JL, Chastre J. Antibiotic stewardship in the intensive care unit. Crit Care. 2014;18(5):480. doi:10.1186/s13054-014-0480-6.
9. Bell BG, Schellevis F, Stobberingh E, Goossens H, Pringle M. A systematic review and meta-analysis of the effects of antibiotic consumption on antibiotic resistance. BMC Infect Dis. 2014;14:13. doi:10.1186/1471-2334-14-13.
10. Schuts EC, Hulscher MEJL, Mouton JW, Verduin CM, Stuart JWTC, Overdiek HWPM, et al. Current evidence on hospital antimicrobial stewardship objectives: a systematic review and meta-analysis. Lancet Infect Dis. 2016;16(7):847-856. doi:10.1016/S1473-3099(16)00065-7.
11. Baur D, Gladstone BP, Burkert F, Carrara E, Foschi F, Dobele S, et al. Effect of antibiotic stewardship on the incidence of infection and colonisation with antibiotic-resistant bacteria and Clostridium difficile infection: a systematic review and meta-analysis. Lancet Infect Dis. 2017;17(9):990-1001. doi:10.1016/S1473-3099(17)30325-0.
12. Karanika S, Paudel S, Grigoras C, Kalbasi A, Mylonakis E. Systematic review and meta-analysis of clinical and economic outcomes from the implementation of hospital-based antimicrobial stewardship programs. Antimicrob Agents Chemother. 2016;60(8):4840-4852. doi:10.1128/AAC.00825-16.
13. Davey P, Marwick CA, Scott CL, Charani E, McNeil K, Brown E, et al. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev. 2017;2(2):CD003543. doi:10.1002/14651858.CD003543.pub4.
14. Tamma PD, Avdic E, Li DX, Dzintars K, Cosgrove SE. Association of adverse events with antibiotic use in hospitalized patients. JAMA Intern Med. 2017;177(9):1308-1315. doi:10.1001/jamainternmed.2017.1938.
15. Feazel LM, Malhotra A, Perencevich EN, Kaboli P, Diekema DJ, Schweizer ML. Effect of antibiotic stewardship programmes on Clostridium difficile incidence: a systematic review and meta-analysis. J Antimicrob Chemother. 2014;69(7):1748-1754. doi:10.1093/jac/dku046.
16. Garnacho-Montero J, Gutierrez-Pizarraya A, Escoresca-Ortega A, Corcia-Palomo Y, Fernandez-Delgado E, Herrera-Melero I, et al. De-escalation of empirical therapy is associated with lower mortality in patients with severe sepsis and septic shock. Intensive Care Med. 2014;40(1):32-40. doi:10.1007/s00134-013-3077-7.
17. Leone M, Bechis C, Baumstarck K, Lefrant JY, Albanese J, Jaber S, et al. De-escalation versus continuation of empirical antimicrobial treatment in severe sepsis: a multicenter non-blinded randomized noninferiority trial. Intensive Care Med. 2014;40(10):1399-1408. doi:10.1007/s00134-014-3411-8.
18. Bouadma L, Luyt CE, Tubach F, Cracco C, Alvarez A, Schwebel C, et al. Use of procalcitonin to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet. 2010;375(9713):463-474. doi:10.1016/S0140-6736(09)61879-1.
19. de Jong E, van Oers JA, Beishuizen A, Vos P, Vermeijden WJ, Haas LE, et al. Efficacy and safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill patients: a randomised, controlled, open-label trial. Lancet Infect Dis. 2016;16(7):819-827. doi:10.1016/S1473-3099(16)00053-0.
20. Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47(11):1181-1247. doi:10.1007/s00134-021-06506-y.
21. Kalil AC, Metersky ML, Klompas M, Muscedere J, Sweeney DA, Palmer LB, et al. Management of adults with hospital-acquired and ventilator-associated pneumonia: 2016 clinical practice guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis. 2016;63(5):e61-e111. doi:10.1093/cid/ciw353.