Ciprofloxacin and co-trimoxazole are standard empiric choices internationally. In Pakistan they are 24% and 27% susceptible. Here is what to use instead, and the trap that comes with it.
The trap this page exists for. The two oral agents that still work in Pakistan — nitrofurantoin and fosfomycin — are exactly the two that cannot treat pyelonephritis, because neither reaches renal tissue. Misclassifying the patient does not pick a weaker drug; it picks one that cannot work at all.
Every recommendation here carries a divergence badge, not an evidence tier.Same as international guidelinePakistan-adaptedLocal data onlyNational guideline outdated
This section does not restate ADA, NICE or IDSA — they do that better. What it shows is where Pakistani practice has to diverge from them, and why.
Decision-support only, and read this first: If your hospital produces an antibiogram, use it and ignore this page for empiric choice. If it does not, this is the argument for producing one — and the two series here differ enough that regional data genuinely changes the answer.
✓ Clinically reviewed by Dr Soban Bin Salman Meer, MBBS · 05 Aug 2026.
Step 1 · Is this actually an infection?
Stewardship first
First: do not treat asymptomatic bacteriuria Same as international guideline
A positive urine culture is not a urinary tract infection. Treating asymptomatic bacteriuria does not prevent symptomatic infection, does not improve outcomes, and drives exactly the resistance shown on this page. In a setting where 81.6% of Enterobacteriaceae in one series were ESBL producers, this is the single highest-yield thing to stop doing.
Only two exceptions:
Pregnancy — screen and treat, because of the risk of pyelonephritis and preterm birth.
Before a urological procedure expected to breach the mucosa.
This is unchanged from international guidance. It is included because it is widely ignored, not because Pakistan differs.
What actually suggests infection
Dysuria, urinary frequency and urgency — the combination is what matters, not a single symptom.
Suprapubic pain or tenderness.
New or worsening incontinence, or a change in urine character, in an older patient.
Loin pain, fever above 38 °C, rigors or vomiting suggest the infection is no longer confined to the bladder.
Not a reason to treat. Cloudy or strong-smelling urine on its own is not a reason to treat. Neither is a positive dipstick in a patient with no urinary symptoms.
Step 2 · Which of the three is this?
Classification drives everything downstream
Which of the three is this? The answer decides whether your drug can work at all
Step 3 · What actually works here
Published Pakistani susceptibility data
Your own hospital antibiogram beats every number on this page. These are two published Pakistani series from different regions and periods. They agree closely on which agents still work and disagree sharply on how badly the others have failed — co-amoxiclav retained 47% susceptibility in Islamabad and none at all in Dera Ismail Khan. Treat these as the shape of the problem when you have nothing local, not as a national standard.
Islamabad, community-acquired uncomplicated UTI Local data only
Retrospective analysis across tertiary care centres, 12,000 samples. Uropathogenic E. coli from uncomplicated UTI in patients presenting from the community. Compared 2005–2010 with 2016–May 2021; figures below are the 2021 endpoint.
Agent
E. coli susceptible
Fosfomycin
92%
Aminoglycosides
90%
Nitrofurantoin
80%
Co-amoxiclav
47%
Cephalosporins
38%
Co-trimoxazole
27%
Fluoroquinolones
24%
Limitations: The authors note the results are not generalisable because the study covers only uncomplicated UTI, and call for mass-level surveys.
Dera Ismail Khan, mixed urinary isolates Local data only
Cross-sectional study, Gomal Medical College and Jamil Medical Lab, January 2022 to November 2024. 610 urine specimens, 260 positive (42.6%). 88% of patients from rural areas. Figures below are for E. coli (n = 155).
Agent
E. coli susceptible
Imipenem
96%
Fosfomycin
87%
Nitrofurantoin
82%
Amikacin
72%
Ciprofloxacin
28%
Ceftriaxone
20%
Ampicillin
5%
Co-amoxiclav
0%
Limitations: Cross-sectional, so it cannot show trend. No molecular analysis of resistance genes. The authors state findings are limited to the region. Some results were inconsistent with CLSI expectations — for example Proteus mirabilis appearing nitrofurantoin-sensitive despite intrinsic resistance.
What grows
Organism
Share of positive cultures
Escherichia coli
59.6%
Enterococcus spp.
11.5%
Klebsiella pneumoniae
10.4%
Proteus mirabilis
6.2%
Pseudomonas aeruginosa
5.8%
Staphylococcus saprophyticus
5%
Staphylococcus aureus
1.5%
ESBL. In the Dera Ismail Khan series, 81.6% of Enterobacteriaceae were ESBL producers. That is the single number that should change how you read every cephalosporin result below.
What this adds up to
Fosfomycin and nitrofurantoin held up in both series — 92% and 80% in Islamabad, 87% and 82% in Dera Ismail Khan. They are the two oral agents you can still start empirically for cystitis.
Fluoroquinolones failed in both — 24% and 28%. Ciprofloxacin is not an empiric option for uncomplicated cystitis in Pakistan.
Co-trimoxazole failed at 27%. It remains first-line in several international guidelines on the basis of local susceptibility above 80%; that condition is not met here.
Cephalosporins and co-amoxiclav are unreliable and regionally variable — 38% and 47% in Islamabad, 20% and 0% in Dera Ismail Khan.
Step 4 · What to start
Driven by your choice in step 2
Step 5 · Before you write the prescription
Three things worth pausing on
Nitrofurantoin and G6PD deficiency — softer than you were taught Pakistan-adapted
The classic teaching is that nitrofurantoin is contraindicated in G6PD deficiency. A safety review of 318 reported episodes of nitrofurantoin-associated haemolytic anaemia found only 42 (13%) had confirmed or highly probable G6PD deficiency, and estimated the risk at roughly 1.3 cases per million exposures across about 245 million documented exposures. Early erythrocyte-survival studies suggest it is less likely to cause oxidant haemolysis than primaquine.
In practice. The review's recommendation is that a total daily dose of 200 mg for a short 3–5 day course may be used without G6PD screening, provided the patient is warned about the signs of haemolysis — dark urine in particular — and told to stop and seek review if they appear.
Labelling genuinely differs by country: the EMA lists a contraindication, the FDA provides only pharmacogenomic information. If your local formulary contraindicates it, follow your formulary. But do not discard one of only two working oral agents on the strength of the classic teaching alone.
Why it is flagged: This matters more in Pakistan than in the source guidelines, because G6PD deficiency is common here and nitrofurantoin is one of only two oral agents local resistance still permits.
Nitrofurantoin and renal function — the threshold has moved, and sources disagree Same as international guideline
Nitrofurantoin needs adequate glomerular filtration to reach therapeutic concentrations in urine. The threshold has been revised repeatedly: the American Geriatrics Society Beers criteria moved from avoiding it below a creatinine clearance of 60 to below 30 in their 2015 update, on the basis of two retrospective studies, while other guidance advises avoiding it below an eGFR of 45 and requires no dose adjustment between 45 and 59.
In practice. Below an eGFR of 30 there is broad agreement not to use it. Between 30 and 45 the sources genuinely differ — decide with your own formulary and the alternatives actually available, rather than assuming a single number is settled.
Why it is flagged: Not a Pakistan-specific issue, but it is contested internationally and worth stating as contested rather than picking a number.
Fosfomycin is a single dose, and that is a feature Same as international guideline
Fosfomycin trometamol is given as one 3 g oral dose for uncomplicated cystitis. In a head-to-head trial, clinical resolution at day 28 was 70% with nitrofurantoin and 58% with fosfomycin — so nitrofurantoin performed better on that endpoint.
In practice. Where a patient is unlikely to complete a five-day course or to return, a single observed dose that works most of the time may still beat a better drug that is abandoned on day two.
Step 6 · Follow-up and what protects the last two agents
Stewardship
No test of cure is needed in an uncomplicated cystitis that resolves. Repeat culture in pregnancy, in complicated infection, and whenever symptoms persist or recur.
Failure to respond within 48–72 hours means one of three things: the organism is resistant, the infection is not confined to the bladder, or there is obstruction. Do not simply change antibiotic without asking which.
Recurrent infection in a man, or in a woman with stones or a structural abnormality, needs urological assessment rather than another course.
The diabetes overlap is not incidental Local data only
In the Dera Ismail Khan series, 68% of patients with a positive urine culture had a history of diabetes. Pakistan has the highest national prevalence of type 2 diabetes in the world. Poor glycaemic control both predisposes to infection and places the patient in the complicated category — so a UTI is a reasonable prompt to check an HbA1c in someone not already known to be diabetic.
What preserves the two agents that still work
Do not treat asymptomatic bacteriuria.
Do not use fluoroquinolones or cephalosporins for uncomplicated cystitis simply because they are available.
Send cultures in complicated infection and pyelonephritis so that carbapenems can be de-escalated rather than continued by default.
Push for a hospital antibiogram. Every number on this page is a substitute for one you should have.
Common questions
Which antibiotic should I use for a simple UTI in Pakistan?
Nitrofurantoin or fosfomycin. In two published Pakistani series, nitrofurantoin retained 80% and 82% susceptibility and fosfomycin 92% and 87%. They are the only oral agents that held up in both.
Can I still use ciprofloxacin for a UTI?
Not empirically for uncomplicated cystitis. Fluoroquinolone susceptibility was 24% in the Islamabad series and 28% in Dera Ismail Khan. That is close to a coin flip that loses, and it spends a drug class that matters elsewhere.
What about co-trimoxazole?
Susceptibility was 27%. International guidelines make co-trimoxazole first-line only where local susceptibility exceeds roughly 80%. Pakistan does not meet that condition.
Why can I not use nitrofurantoin for pyelonephritis?
Nitrofurantoin does not achieve adequate concentrations in blood or deep renal tissue, so it cannot treat a kidney infection regardless of what the susceptibility report says. Fosfomycin is likewise not a pyelonephritis drug. This is the trap in Pakistani practice: the two agents that still work for cystitis are exactly the two that cannot treat the kidney.
Is nitrofurantoin contraindicated in G6PD deficiency?
The classic teaching says so, but a safety review found only 42 of 318 reported cases of nitrofurantoin-associated haemolytic anaemia had confirmed or probable G6PD deficiency, and estimated the risk at about 1.3 cases per million exposures. It recommends short 3–5 day courses at 200 mg daily may be given without G6PD screening, provided the patient is warned about dark urine and told to stop if it occurs. Labelling differs by country, so follow your formulary — but do not discard the drug on the classic teaching alone.
Should I treat a positive urine culture in someone with no symptoms?
No, except in pregnancy or before a urological procedure that will breach the mucosa. Treating asymptomatic bacteriuria does not improve outcomes and drives the resistance that has already removed most of the options here.
How reliable are these resistance figures for my hospital?
Less reliable than your own antibiogram, which should always take precedence. The two series agree closely on which agents still work and disagree sharply on the others — co-amoxiclav retained 47% susceptibility in Islamabad and none at all in Dera Ismail Khan. Treat them as the shape of the problem when you have nothing local.
Why does the ESBL rate matter so much?
In the Dera Ismail Khan series 81.6% of Enterobacteriaceae were ESBL producers. That makes empiric cephalosporins unreliable and puts carbapenems under real pressure — which is why sending cultures so you can de-escalate matters more here than in the settings these guidelines were written for.
Clinically reviewed by Dr Soban Bin Salman Meer, MBBS
MBBS — Postgraduate Trainee in the Department of Urology at the Pakistan Institute of Medical Sciences (PIMS), Islamabad. Graduated from Federal Medical & Dental College, Islamabad.
Sources
NO PAKISTANI NATIONAL UTI GUIDELINE EXISTS as far as we can establish. The clinical structure below follows standard international practice; the empiric agent choice is driven by published Pakistani susceptibility data, cited so you can weigh it yourself.
Twelve-year trend of Escherichia coli antibiotic resistance in the Islamabad population — 12,000 samples, tertiary care centres, community-acquired uncomplicated UTI, 2005–2010 versus 2016–2021.
Bacterial profiling and antibiotic resistance patterns in urinary tract infections: a microbiological analysis from Dera Ismail Khan, Pakistan. BMC Infectious Diseases, cross-sectional, January 2022 to November 2024, 610 specimens, 260 positive.
Nitrofurantoin and glucose-6-phosphate dehydrogenase deficiency: a safety review — the source of the 1.3 per million risk estimate and the recommendation that short courses may be given without G6PD screening.
Nitrofurantoin renal thresholds: American Geriatrics Society Beers Criteria 2015 update (creatinine clearance below 30) versus guidance advising avoidance below an eGFR of 45.
Background reading — why ciprofloxacin is a one-in-four bet here, the pyelonephritis trap, and the G6PD rule that is softer than you were taught: UTI in Pakistan.