For Clinicians | Psychiatric Medications in a Disaster

Why Stopping and Restarting Are Two Different Hazards

By Dr. Jamie Wilkey, PharmD — Director of Clinical Strategy, Jase
Medically reviewed and edited by Kristen Carpenter, PA-C

Our last article ranked a patient’s Rx medicine cabinet so you could tell a patient which of their prescriptions genuinely can’t wait and are the most important. One row on that table needed a lot more room than we gave it. Disaster planning almost always lands on insulin and blood pressure medication, and that’s not anybody’s oversight. It’s where the evacuee data pointed, so it’s where the guidance went. We went looking for mental health medications on the standard kit lists and came up empty.

So this write up is only about mental health medications. Stopping them is one hazard and restarting them is a separate one. We’re covering two specific drugs, clozapine and lithium, and one whole class, the benzodiazepines, because all three fail in a different direction when access is disrupted with the rules of thumb that hold across the category.

Clozapine: the number of days decides the dose

When a patient on clozapine has had any gap in dosing, the first thing to establish is the date of the last dose.¹ The length of the gap is what sets the restart dose. Their maintenance dose is not the default answer.

The mechanism is loss of tolerance to clozapine’s hypotensive effect.² Current labeling puts orthostatic hypotension, bradycardia, syncope and cardiac arrest among the things that can happen “when restarting patients who have had even a brief interruption in treatment,” at doses as low as 12.5 mg a day.² The same labeling names dehydration as a predisposing condition, which is the state a lot of these patients are in during a natural disaster.

US labeling sets the restart by days missed:²

• One day missed: resume at 40% to 50% of the previous dose
• Two days missed: resume at roughly 25% of the previous dose
• Anything longer: restart at 12.5 mg once or twice daily, then climb back faster than a new start

Clozapine can drive down neutrophils, the white cells that fight bacterial infection, so these patients have to get regular blood draws to check the absolute neutrophil count (ANC). If that count falls far enough, an ordinary infection can turn life-threatening. A break under 30 days doesn’t change the ANC monitoring schedule. At 30 days or more, monitoring reverts to the weekly-for-six-months schedule of a new start.² A one-week outage changes the dose and leaves the bloodwork alone. Until February 2025, the Clozapine REMS made a pharmacy verify an ANC before dispensing, so the monthly refill didn’t go out without a current lab value on file, and a patient who had drifted out of monitoring hit a hard stop at the counter.³ That backstop is gone, so the extra eye on a clozapine patient has to be yours now.

Benzodiazepines: write the plan down while you still can

For any patient on a daily benzodiazepine, the contingency plan needs to be documented while they can still reach you. Labeling is really clear and tells us why: abrupt discontinuation or rapid dose reduction after continued use “may precipitate acute withdrawal reactions, which can be life-threatening,” seizures among them.⁴ The 2025 joint tapering guideline, ten societies including ASAM and the APA, says anyone past a month of use shouldn’t stop abruptly and should taper under supervision.⁵

Withdrawal symptoms start 2 to 10 days after the last dose and can run for weeks.⁶ This is exactly why it is important to have a plan ahead of time for these medications. Withdrawal in this class needs active treatment, and a disaster is when active treatment is hardest to reach.⁶ A patient in withdrawal during an evacuation is agitated, tremulous, and tachycardic, in a setting with nothing to manage it with. 

Controlled-substance prescribing follows the same federal and state rules during a declared emergency so there isn’t disaster wiggle room here. The built-in emergency provisions are narrow, and some states bar writing for a patient the provider hasn’t examined unless that rule gets waived.⁷ 

The biggest reason patients lose access to key medications like this is structural. That means no money, no services running, or no way to get there. After Katrina, that was the reason for 74% to 84% of the people who were already in mental health treatment and then cut back or stopped.⁸ So these are the patients whose plan you prioritize. Getting through a natural disaster on a stable regimen is hard enough. Doing it in withdrawal is worse, and that’s the part planning ahead can actually change.

Lithium: the risk shows up without a missed dose

The conversation for a lithium patient is particularly interesting because in a disaster it isn’t about the medication itself that we have to worry about so much. It’s more about what the patient taking lithium is exposed to: heat, fluids, salt, and what pain medication they are using.

The kidney handles lithium a lot like sodium, so sodium or volume depletion drives reabsorption up and excretion down.⁹ Serum levels climb on an unchanged dose. Dehydration, vomiting, diarrhea, fever, and a low-sodium intake all contribute. So does a week of heat with no air conditioning, which is how a power outage affects a lithium level.

Then there’s the part that comes from the medicine cabinet instead of the weather. NSAIDs and ACE inhibitors both raise lithium levels by changing how the kidney handles sodium.⁹ A patient with a sprained ankle and a bottle of ibuprofen in a house with no power has assembled a lithium problem out of ordinary decisions.

The counseling point is early toxicity: a new tremor, GI upset, feeling unsteady.⁹ Those are vauge-ish symptoms that a patient may very well shake off as just being part of living through a natural disaster and miss them as warning signs. If these warning signs manifest then that’s a call to the prescriber, not a dose the patient adjusts on their own.

Everything else on the list

Clozapine, benzodiazepines, and lithium are the high-alert three. Most of your patients are on something else and the risk on those is different. A gap produces withdrawal symptoms, and over a longer stretch relapse, rather than the acute physiologic events the first three carry.

Two things predict how hard your patients will be hit. The first is half-life, which is just how long the drug is still doing its job after the last dose. Nobody has these memorized, and looking one up takes awhile, so here they are.¹⁰

The second is how long they’ve been on it. Someone a few weeks into an SSRI is at low risk of serious withdrawal. Someone eight years in certainly is at a much, much higher risk of withdrawal.¹¹

Half-life is a first pass, not a verdict. Sertraline’s half-life sits above paroxetine’s, and it made almost no difference: after 5 to 8 days off, withdrawal showed up in 60% of patients taking sertraline and 66% on paroxetine, against 14% on fluoxetine.¹² The table sorts the extremes reliably. In the middle it’s a rough guide.

Oral antipsychotics and mood stabilizers don’t sort this way. Their gap risk is relapse over weeks to months rather than withdrawal over days, so half-life doesn’t predict it.

Restarting is simpler here than with clozapine. Nothing has lost tolerance, so a short gap generally resumes at the prior dose. A long one doesn’t. Reinstating a drug months after stopping it can behave unpredictably, including symptoms worsening rather than settling.¹¹

One thing not to counsel: stretching the interval. Spacing doses to every other day is the intuitive way to make what’s left go further, and for most of this list it’s the wrong move. Because so many of these half-lives run under a day, every-second-day dosing swings concentrations up and down and can precipitate severe withdrawal, which is the outcome the patient was trying to avoid.¹¹ If someone is going to run short, the prescriber reduces the dose rather than widening the interval.

Two halves of preparation

We want patients protected in a disaster, and the ones on mental health medications have been left out of many disaster preparation conversations.

Appropriate medical preparation here comes in two halves. The first is clinical, and it’s the whole article above: knowing which of your patients has no buffer, knowing that the restart is a prescriber decision rather than the patient’s, and having that decision on paper before anything happens. That first half is not a quantity, and for this class it can’t be, since controlled substances like benzodiazepines have so much red tape around dispensing them.

The second half is boring, and it’s the one that gets skipped. The formal, active, updated medication list that travels with the patient. In one Japanese flood study, 48% of evacuees left their medication behind and 88% left their prescription records behind.¹³ People remember the pills more reliably than the paper, and the paper is what lets any pharmacist or any prescriber anywhere pick up the thread.

None of this replaces primary care. It’s the layer that helps prepare a patient well with your help ahead of if/when primary care isn’t reachable.

Jase Daily: extends the list, not the benzodiazepines

Jase Daily extends a stable, existing prescription up to a 12-month supply for a patient your practice has seen within the past 6 months. That covers lithium and most of the SSRIs and SNRIs above. It doesn’t cover benzodiazepines. Jase’s pharmacy doesn’t dispense controlled substances, so a benzodiazepine patient’s continuity still runs through the written plan, not a supply workaround.

Sources

1. ACLP How To Guide: Clozapine Consultation. Academy of Consultation-Liaison Psychiatry, 2024. Directs clinicians to determine the patient’s last clozapine dose as soon as possible, and gives a re-initiation schedule that differs from US labeling: no re-titration up to 48 hours, half the previous total daily dose then 75% then full dose over three days for a 48 to 72 hour break, 12.5 mg once or twice daily from 72 hours to one week, and titration as a new patient beyond a week. http://clpsychiatry.org/wp-content/uploads/ACLP-How-To-Guide-Clozapine-2025.pdf 
2. Clozaril (clozapine). FDA prescribing information, revised June 2025. Identical language across Clozaril, clozapine tablets, and clozapine orally disintegrating tablets. Warnings state that orthostatic hypotension, bradycardia, syncope, and cardiac arrest have occurred, and that these reactions can occur with the first dose, at doses as low as 12.5 mg per day, or when restarting patients who have had even a brief interruption in treatment; dehydration is named among predisposing conditions. Section 2.6 sets the restart at 40% to 50% of the previous dosage after one missed day, roughly 25% after two, and 12.5 mg once or twice daily for longer gaps. ANC testing continues at the previous frequency for interruptions under 30 days and reverts to the initiating-treatment schedule, weekly for months 1 through 6, at 30 days or more. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=90876802-0e3a-44c9-9ff7-1754dfbe736a 
3. Clozapine REMS has ended. National Community Pharmacists Association, 2025. The program ended effective February 24, 2025; pharmacies no longer need ANC results before dispensing, and FDA continues to recommend prescribers monitor ANC at labeled frequencies. The current labeling contains no REMS references while retaining the ANC monitoring table. https://ncpa.org/newsroom/qam/2025/03/31/clozapine-rems-has-ended 
4. Xanax (alprazolam). FDA prescribing information, revised April 2024. Boxed Warning states that continued use may lead to clinically significant physical dependence, and that abrupt discontinuation or rapid dosage reduction after continued use may precipitate acute withdrawal reactions, which can be life-threatening; section 5.3 names seizures as an example. Dependence risk is greater above 4 mg daily and beyond 12 weeks but exists after shorter use at lower doses. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=bb5cf72e-3a41-41a8-9a38-80f974eca33f 
5. Brunner E, et al. Joint Clinical Practice Guideline on Benzodiazepine Tapering: Considerations When Risks Outweigh Benefits. Journal of General Internal Medicine, 2025;40(12):2814-2859. Sponsored by ten organizations including ASAM, the AAFP, and the APA. Patients taking benzodiazepines longer than a month should not abruptly discontinue and should taper under clinical supervision. https://www.asam.org/quality-care/clinical-guidelines/benzodiazepine-tapering 
6. Regina AC, Gokarakonda SB, Attia FN. Withdrawal Syndromes. StatPearls, NCBI Bookshelf. Benzodiazepine withdrawal develops 2 to 10 days after discontinuation and can last weeks; seizures are named among the physical signs; benzodiazepine and barbiturate withdrawal are described as potentially life-threatening and requiring aggressive management. https://www.ncbi.nlm.nih.gov/books/NBK459239/ 
7. Rutkow L, et al. Prescribing authority during emergencies: challenges for mental health care providers. Journal of Legal Medicine, 2011. Providers must follow the federal and state laws regulating prescribing during an emergency, the emergency provisions in those laws are limited, and some states do not permit prescribing for a patient the provider has not examined unless that provision is waived during a declared emergency. https://pmc.ncbi.nlm.nih.gov/articles/PMC3229266/ 
8. Wang PS, et al. Disruption of existing mental health treatments and failure to initiate new treatments after Hurricane Katrina. American Journal of Psychiatry, 2008;165(1):34-41. Among survivors with a pre-hurricane mental disorder, 12.7% terminated and 10.2% reduced treatment, 22.9% combined. Barriers involving enabling factors, meaning financial barriers and unavailability of services plus transportation problems outside the New Orleans metro subsample, were the most commonly reported reasons for reducing or terminating treatment, at 84.0% and 74.0% across the two subsamples. https://pmc.ncbi.nlm.nih.gov/articles/PMC2248271/ 
9. Slubicki M, et al. Lithium Toxicity. StatPearls, NCBI Bookshelf. Because the kidney handles lithium similarly to sodium, sodium or volume depletion increases renal lithium reabsorption and decreases excretion; precipitants include dehydration, vomiting, diarrhea, fever, and low-sodium diets. NSAIDs and ACE inhibitors alter renal hemodynamics or sodium handling, increasing reabsorption or reducing clearance. https://www.ncbi.nlm.nih.gov/books/NBK499992/ 
10. Keks N, Hope J, Keogh S. Switching and stopping antidepressants. Australian Prescriber, 2016;39(3):76-83, which tabulates elimination half-lives: fluoxetine 4 to 16 days including norfluoxetine, vortioxetine 2.4 to 2.8 days, citalopram and escitalopram 1.5 days, sertraline 1.1 to 1.3 days, paroxetine 1.0 day, mirtazapine 0.8 to 1.6 days, amitriptyline 0.2 to 1.9 days, venlafaxine 0.6 days including desvenlafaxine, fluvoxamine 0.6 days, duloxetine 0.5 days, desvenlafaxine 0.4 days. https://pmc.ncbi.nlm.nih.gov/articles/PMC4919171/ Bupropion is not in that table: 21 hours, with erythrohydrobupropion about 33 and threohydrobupropion about 37 hours, from Wellbutrin XL prescribing information, https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=a435da9d-f6e8-4ddc-897d-8cd2bf777b21 Trazodone is biphasic, 3 to 6 hours initial and 5 to 9 hours terminal, from Cuomo A, et al., A narrative review on trazodone as a multimodal and multifunctional antidepressant, Annals of General Psychiatry, 2026;25:32, https://pmc.ncbi.nlm.nih.gov/articles/PMC13112903/ 
11. Therapeutics Letter 157: How to stop antidepressants. Therapeutics Initiative, University of British Columbia, 2025. Because the elimination half-life of many antidepressants is under a day, every-second-day dosing can produce wide fluctuations in drug concentrations and precipitate severe withdrawal effects. Patients on an antidepressant for only several weeks are at low risk of serious withdrawal compared with those with years of exposure, and resuming a withdrawn drug at a low dose months after stopping can have unpredictable effects including paradoxical worsening. https://www.ncbi.nlm.nih.gov/books/NBK618767/ The alternate-day modelling it draws on is O’Neill JR, et al., Alternate-day dosing to taper antidepressants risks severe withdrawal effects: an in silico analysis, Journal of Affective Disorders, 2026;392:120084, https://pubmed.ncbi.nlm.nih.gov/40848773/ 
12. Therapeutics Letter 156: Antidepressant withdrawal syndrome, update. Therapeutics Initiative, University of British Columbia, 2025. After 5 to 8 days of discontinuation, withdrawal occurred in 14% of patients taking fluoxetine against 60% taking sertraline and 66% taking paroxetine. Paroxetine, duloxetine, and venlafaxine/desvenlafaxine are disproportionately represented in WHO withdrawal reporting. https://www.ncbi.nlm.nih.gov/books/NBK618768/ 
13. Ochi S, et al. Disaster-Driven Evacuation and Medication Loss: a Systematic Literature Review. PLoS Currents Disasters, 2014. Seventy articles covering 27 events across 9 countries. Following Japan floods, 48% of evacuees left their medication behind and 88% left their prescription records behind. The 5.7x prepared-emergency-pack finding also comes from a single Japanese flood study inside this review rather than pooled across its 27 events. https://pmc.ncbi.nlm.nih.gov/articles/PMC4169391/ 

For Clinicians | Which Medications Can’t Be Missed

How to Help Patients Rank their Medication List Before a Disruption

By Dr. Jamie Wilkey, PharmD — Director of Clinical Strategy, Jase
Medically reviewed and edited by Kristen Carpenter, PA-C

September is National Preparedness Month and that’s a big deal for us over here at Jase. When it comes to medications, your patients get the same advice they get every year: keep some extra medication on hand. But between all the prescription medications they take and the OTC products and supplements, that one line of advice quickly grows confusing and unmanageable. Nobody has ever published a ranking of which medications matter most, so patients sort it themselves.

Why there isn’t a ranking

One team went through 186 academic articles and 252 web pages of preparedness guidance for chronically ill patients, spanning cancer, cardiovascular disease, diabetes, chronic respiratory disease, and kidney disease. Recommended amounts on hand for the medications that treat these conditions ran anywhere from having a cushion of 3 days to a full month depending on which source you opened. Most of the recommendations traced back to expert opinion rather than evidence. None of it answers which medications actually tolerate an interruption. The review catalogues how much to have and never asks which one matters most.¹

To make matters more opaque, the literature it covered ends in 2014. The gap is still open today. So when a patient asks us how much to keep on hand, we don’t have anything more than generalities to give them.

How to rank a list

Here’s the 4 questions we use: 

1. How sick does this patient get without this drug?
2. How fast does that happen?
3. How likely is this one to be short when everyone needs it?
4. Can this particular patient actually get to a pharmacy in a bad week? 

Run a list through those and the order falls out. It’s not always what the patient expects. The table below answers the first two for the most common classes. Those are drug properties. The last two are yours, because only you know this patient.

Three exceptions

The table above is great for a quick scan. Steroids, clonidine and seizure meds need more context than the table can hold.

Ask every patient on long-term steroids what they do when they get sick. If they don’t have an answer, that’s your time to shine in educating them. Physiologic stress raises the requirement, so the sick-day rule is to double the oral dose at a fever above 100.4F and triple it above 102.2F.12 Omission during stress is what precipitates adrenal crisis. A patient rationing their bottle is under stress and cutting the dose at the same time. Something to watch for sure.

Check whether your clonidine patients are also on a beta-blocker. Stopping clonidine suddenly can push blood pressure higher than it was before treatment, and being on a beta-blocker too makes that worse.4 That’s why an elective taper stops the beta-blocker several days first. Patients on both would need top priority attention if access to those medications were disrupted.

And now for epilepsy. Sustained nonadherence tracks with worse seizure control and higher mortality.¹⁰  But a prospective study published this year in drug-resistant epilepsy found that occasional missed doses did not acutely raise seizure risk.11 So they sit in the middle of the ranking, not at the top. A missed dose is not the emergency. Multiple missed doses are. 

You intuitively know which of these goes first. The patient doesn’t, not off the top of their head. So the next time their list is in front of you, help them sort it into what’s critical and what has some flexibility.

Where we sit, and what to take away

This is where Jase specializes: the undefined middle of appropriate medical preparation. We help patients get ahead of a disruption before it throws off a critical regimen, with a licensed provider reviewing every request. Disasters hit chronic conditions unevenly, and we help sort out what comes first. A complement to primary care, never a replacement for it.

The ranking is the thing to walk away with, and it’s an order rather than an amount. Every list has one drug that is most critical, and most patients aren’t clear exactly what that one is.

And now before you go you should know we still have one class that is so important that it needs its own article: psychiatric medications. Stopping them is one hazard and restarting them is another. So our next article will be a deep dive on mental health medications in a disaster.

Sources

1. Tomio J, Sato H. Emergency and disaster preparedness for chronically ill patients: a review of recommendations. Open Access Emergency Medicine, 2014. 186 academic articles and 252 web pages screened; recommended supply ranged from 3 days to 1 month; the authors state that most of the recommendations have not been validated scientifically and that a large number rest on anecdotal evidence or expert opinion. https://pmc.ncbi.nlm.nih.gov/articles/PMC4753992/
2. Administration and Pharmacokinetics of Levothyroxine, in 70 Years of Levothyroxine. NCBI Bookshelf. Elimination half-life averages roughly 6 to 7 days in euthyroid adults and about 7.5 days in treated hypothyroid patients, which is what supports once-daily dosing. https://www.ncbi.nlm.nih.gov/books/NBK585644/
3. Amlodipine besylate (NORVASC). FDA prescribing information via DailyMed. Terminal elimination half-life of about 30 to 50 hours, from which a return toward baseline over several days follows. The label separately notes that amlodipine gives no protection against the dangers of abrupt beta-blocker withdrawal, which is the distinction this row turns on. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=b00570ff-4081-4514-b45b-eb6f753450f5
4. Clonidine. StatPearls, NCBI Bookshelf: abrupt discontinuation may result in rebound hypertension, the risk is higher with concurrent beta-blockade through unopposed alpha-adrenergic activity, and beta-blockers are withdrawn several days before tapering clonidine. https://www.ncbi.nlm.nih.gov/books/NBK459124/ Catapres (clonidine hydrochloride) FDA label: sudden cessation has produced nervousness, agitation, headache and tremor with a rapid rise in blood pressure and elevated plasma catecholamines, and rare instances of hypertensive encephalopathy, cerebrovascular accident and death have been reported after withdrawal. https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/017407s037lbl.pdf
5. Asthma: Rank MA, et al. The risk of asthma exacerbation after stopping low-dose inhaled corticosteroids: a systematic review and meta-analysis of randomized controlled trials. Journal of Allergy and Clinical Immunology, 2013. Patients with well-controlled asthma who stop regular low-dose inhaled corticosteroids have increased exacerbation risk compared with those who continue. https://pubmed.ncbi.nlm.nih.gov/23321206/ COPD: Magnussen H, et al. Withdrawal of inhaled glucocorticoids and exacerbations of COPD. New England Journal of Medicine, 2014;371(14):1285-1294. 2,485 patients with a history of COPD exacerbation on triple therapy; stepwise fluticasone withdrawal over 12 weeks was noninferior for moderate or severe exacerbations, with a greater decline in lung function during the final withdrawal step. https://pubmed.ncbi.nlm.nih.gov/25196117/
6. Burge MR, Schmitz-Fiorentino K, Fischette C, Qualls CR, Schade DS. A prospective trial of risk factors for sulfonylurea-induced hypoglycemia in type 2 diabetes mellitus. JAMA, 1998. No hypoglycemia occurred across 156 fasting studies in elderly patients receiving maximum sulfonylurea doses. The retrospective literature reaches the opposite conclusion, with omission of one or more meals implicated in 80% of severe drug-induced hypoglycemia admissions; risk concentrates in renal or hepatic impairment, alcohol use, and polypharmacy rather than in fasting alone. https://jamanetwork.com/journals/jama/fullarticle/1150338
7. Reimer C, Sondergaard B, Hilsted L, Bytzer P. Proton-pump inhibitor therapy induces acid-related symptoms in healthy volunteers after withdrawal of therapy. Gastroenterology, 2009;137(1):80-87. Randomized double-blind placebo-controlled trial in 120 healthy volunteers given 8 weeks of esomeprazole 40 mg followed by 4 weeks of placebo; clinically relevant heartburn, acid regurgitation, and dyspepsia appeared after withdrawal and persisted at least 2 weeks. Conducted in healthy volunteers, not chronic PPI patients. https://pubmed.ncbi.nlm.nih.gov/19362552/
8. Ketones and Insulin Pumps. UCSF Diabetes Teaching Center. Pump therapy carries higher ketone risk because long-acting insulins are not used and rapid-acting insulin is delivered in very small amounts; check glucose and ketones if basal delivery is interrupted for more than an hour. https://diabetesteachingcenter.ucsf.edu/content/ketones-and-insulin-pumps
9. Ferri N, Colombo E, Tenconi M, Baldessin L, Corsini A. Drug-Drug Interactions of Direct Oral Anticoagulants (DOACs): From Pharmacological to Clinical Practice. Pharmaceutics, 2022;14(6):1120. Elimination half-lives: rivaroxaban 5 to 9 hours in healthy adults, edoxaban 8 to 11 hours, apixaban 8 to 15 hours, dabigatran 12 to 17 hours. https://pmc.ncbi.nlm.nih.gov/articles/PMC9229376/
10. Faught E, et al. Nonadherence to antiepileptic drugs and increased mortality: findings from the RANSOM Study. Neurology, 2008. Medicaid claims analysis of 33,658 adults with epilepsy across 388,564 treated quarters, 26% of them nonadherent; nonadherence was associated with more than a threefold increase in mortality, hazard ratio 3.32 (95% CI 3.11 to 3.54). https://pubmed.ncbi.nlm.nih.gov/18565827/
11. Goldenholz DM, et al. Does Missing Medication Acutely Change Seizure Risk? A Prospective Study. Annals of Neurology, 2026. Prospective community-based cohort of 27 adults with drug-resistant epilepsy, defined as three or more seizures per month, followed 10 months each across 7,853 person-days with 93% average adherence; occasional missed doses did not measurably raise immediate seizure risk, while prior seizure history did predict future seizures. https://pmc.ncbi.nlm.nih.gov/articles/PMC12782288/
12. Adrenal Crisis. StatPearls, NCBI Bookshelf. Sick day rules: doubling the oral corticosteroid dose for fever exceeding 100.4°F (38°C) and tripling it for fever exceeding 102.2°F (39°C). https://www.ncbi.nlm.nih.gov/books/NBK499968/

For Clinicians | Purulent vs. Nonpurulent Cellulitis

What an Antibiotic Can’t Do for a Wound

By Dr. Jamie Wilkey, PharmD — Director of Clinical Strategy, Jase
Medically reviewed and edited by Kristen Carpenter, PA-C

Most of what goes wrong with a wound isn’t fixed by an antibiotic. Your patients are working that out on day three, alone, with something left over in the closet. Today we’re covering antibiotic prescribing basics, the things a pill can’t touch, and the presentations that get seen rather than treated.

Purulent and nonpurulent: two kinds of wound trouble

Almost everything that goes wrong with a wound is one of two problems, and they behave nothing alike.

Purulent is a pocket. Pus collects in one place, the body walls it off, and pressure builds inside. You can feel it: a tender lump that gets firmer, then squishy in the middle. Abscess, furuncle, carbuncle, all the same problem at different sizes. It doesn’t spread so much as it grows and tightens, and it keeps going until something opens it, either a blade or the skin giving way on its own. 

Nonpurulent is a spread. There’s no pocket to find. Infection moves outward through the tissue, so the red edge sits somewhere different tomorrow than it did today. That’s cellulitis, and the border is the thing to watch. 

An antibiotic travels in the blood, and blood is the one thing that doesn’t reach the middle of a walled-off pocket. Therefore, a pocket has to get opened for treatment. A spread gets a drug to heal because that drug will actually reach the infection site. Draining an abscess is the treatment, not the prep work before the real treatment.

Which drug does which job

Purulent, once it’s drained, often needs nothing at all. That surprises people, so here it is straight from the guideline: for a simple abscess that’s been opened, systemic antibiotics are unnecessary, even when the organism is MRSA.¹ The drainage did the work. A drug gets added when the patient meets SIRS criteria or has genuinely impaired host defenses, not because a scary name came back on the culture.¹

Nonpurulent is the one that actually needs a drug. There’s nothing to open, so the treatment is systemic and it aims at strep: penicillin VK or cephalexin.¹ Cephalexin is 500 mg every six hours. The twice-a-day scripts that come through for a red leg are underdosed for it. The article reviewer, Kristen, has a family member who is an infectious disease doctor and she always says so many people underdose cephalexin and that is the reason it fails, not because it was “resistant”.

So of the two problems in front of you, the pill is the primary treatment in exactly one. The other one needs a blade.

When MRSA actually enters the picture

Pus means staph, and staph in the community is frequently MRSA. No pus usually means strep, and strep is never MRSA. So MRSA lives in the abscess, and the abscess, once it’s drained, usually needs no antibiotic at all.

This leaves our MRSA treatment reflex pointed at the wrong wound. The red leg with no pocket is where doxycycline gets added in practice, and it’s the presentation where MRSA is least likely to be.

The pair itself isn’t wrong. Cephalexin covers strep and misses MRSA. Doxycycline and trimethoprim-sulfamethoxazole cover MRSA and are shakier on strep. When you truly need both, IDSA’s oral answer is clindamycin alone or a beta-lactam paired with one of those two.¹ A real regimen, but used inappropriately at times when only strep needed covering.

So somebody ran that trial. Five hundred patients with uncomplicated cellulitis, cephalexin plus Bactrim against cephalexin plus placebo. Cure rates came back two points apart, with cephalexin alone on the higher end.² Cephalexin plus Bactrim isn’t a better treatment than cephalexin alone, and the trial’s own authors couldn’t rule out a small benefit they’d want tested again.

Nobody gets to say cephalexin alone wins. But it isn’t the no-brainer it feels like either, and that’s enough to make the next red leg worth a “why are we still doing this?”

MRSA coverage (Bactrim or doxycycline by mouth) belongs in two places: a drained abscess in someone with systemic signs or impaired host defenses, and cellulitis where there’s a real reason to suspect it, like penetrating trauma, known MRSA, or injection drug use.¹

What a pill can’t touch

Tetanus is the big one here. The ten-year vaccine interval is for clean, minor wounds. For a dirty one, and that means a puncture, a crush, a burn, or anything carrying soil or saliva, the interval is five years.³ Somebody eight years out from their last booster who steps on a nail in the garden is due, even though the ten-year rule will tell them they’re fine. Either Td or Tdap works for wound prophylaxis.³

Diabetes and peripheral vascular disease change the math, in two different ways that get collapsed into one. Poor perfusion means the wound closes slowly and infection is harder to clear. Neuropathy is the other problem, and it sits upstream of all of it: an injury goes unnoticed because the pain that normally makes a person look never arrives. By the time the wound is seen, it’s older than the patient’s story suggests.⁴

Then the one that isn’t an antibiotic decision at all. Pain out of proportion to what the skin shows, a border moving while you watch it, systemic toxicity, crepitus or bullae. That’s a necrotizing infection until proven otherwise, and IDSA’s line is prompt surgical consultation:¹ a surgeon tonight rather than a prescription and a recheck in the morning.

The guideline is twelve years old

The guideline I keep quoting is from 2014, and it’s still the current US document. Two randomized trials since then complicate what I told you about drained abscesses. Talan found Bactrim after drainage raised cure from roughly 74% to 80%.⁵ Daum found the same direction in abscesses under 5 cm.⁶ Both landed within three years of the guideline, and the guideline still says what it said.

None of it answers what the patient is actually asking, which is what to do about their own leg on day three.

The missing middle

Right now this is all or none. Either the patient gets to you and you look at the leg, or they’re navigating this on their own (offgrid, trapped by a natural disaster, or in a foreign country to name a few scenarios). Nothing sits in between, where somebody has already told them which version of this kind of infection needs a healthcare professional’s eyes on it.

That in-between is our work, and it has a name: appropriate medical preparation. It isn’t a replacement for primary care, and this article is the argument for why. Half of what we’ve covered is something only a clinician can do.

We’re medical doctors, PAs, and pharmacists, and we spend our time charting the parts of this that the guidelines leave grey. If a patient is pressing you for something to keep on hand and you don’t have the bandwidth for it, send them to us at Jase.com. A licensed provider reviews the request and writes the prescription, and we’ll keep publishing where we draw the lines.

The bottom line

Most of what goes wrong with a wound isn’t fixed by an antibiotic. Some of it gets opened, some of it is a tetanus shot, some of it needs a surgeon tonight, and a good share of it was never an infection at all.

When specialists re-examined patients who’d been diagnosed with cellulitis, 41% had something else, usually stasis dermatitis, eczema, or lymphedema.⁷ That’s trained clinicians, examining the patient in person, getting it wrong four times out of ten. Your patient, looking at their own leg and typing symptoms into a phone, is not going to do better.

Sources

1. Stevens DL, Bisno AL, Chambers HF, et al. Practice Guidelines for the Diagnosis and Management of Skin and Soft Tissue Infections: 2014 Update by the Infectious Diseases Society of America. Clin Infect Dis. 2014;59(2):e10-e52. https://academic.oup.com/cid/article/59/2/e10/2895845 
2. Moran GJ, Krishnadasan A, Mower WR, et al. Effect of Cephalexin Plus Trimethoprim-Sulfamethoxazole vs Cephalexin Alone on Clinical Cure of Uncomplicated Cellulitis: A Randomized Clinical Trial. JAMA. 2017;317(20):2088-2096. https://pmc.ncbi.nlm.nih.gov/articles/PMC5815038/ 
3. Centers for Disease Control and Prevention. Tetanus. Epidemiology and Prevention of Vaccine-Preventable Diseases (Pink Book), chapter 21. https://www.cdc.gov/pinkbook/hcp/table-of-contents/chapter-21-tetanus.html 
4. Senneville É, Albalawi Z, van Asten SA, et al. IWGDF/IDSA Guidelines on the Diagnosis and Treatment of Diabetes-related Foot Infections (IWGDF/IDSA 2023). Clin Infect Dis. 2023. https://pubmed.ncbi.nlm.nih.gov/37779323/ 
5. Talan DA, Mower WR, Krishnadasan A, et al. Trimethoprim-Sulfamethoxazole versus Placebo for Uncomplicated Skin Abscess. N Engl J Med. 2016;374(9):823-832. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4851110/ 
6. Daum RS, Miller LG, Immergluck L, et al. A Placebo-Controlled Trial of Antibiotics for Smaller Skin Abscesses. N Engl J Med. 2017;376(26):2545-2555. https://pmc.ncbi.nlm.nih.gov/articles/PMC6886470/ 
7. Nightingale R, et al. Misdiagnosis of Uncomplicated Cellulitis: a Systematic Review and Meta-analysis. J Gen Intern Med. 2023;38(10):2396-2404. https://link.springer.com/article/10.1007/s11606-023-08229-w