19.2 Pharmacotherapeutic Outcome Monitoring

Official Domain V knowledge includes pharmacotherapeutic intervention and treatment outcomes. Chapter 9 taught you how the drug works and which pair you must not start. Chapters 10–16 taught system-specific regimens. This section is the clock, the lab, and the decision after the first prescription: therapeutic effect, adverse effect, or no effect — and what you do next.

ANCC will hand you a day-11 sertraline visit, a 10-day lisinopril BMP, or a day-3 undrained abscess and ask whether the drug failed. The trap is evaluating before the clock or ignoring the harm sitting next to a prettier disease number.

Three outcome buckets — do not collapse them

Every follow-up visit sorts the drug into one of three buckets. Mixing them is how people get harmed.

“Fair trial” is the entire game. A 10-day SSRI at 10 mg is not a failed antidepressant. A 3-day ACE inhibitor is not a failed antihypertensive. An antibiotic started yesterday is not “resistant.” Conversely, a drug that dropped A1c to 6.3% while producing weekly hypoglycemia is not a therapeutic success — it is an adverse-effect outcome wearing a quality-measure costume (Section 19.4).

Time-to-effect — the clocks ANCC expects

Do not evaluate a drug before it could have worked. These intervals are the ones family primary care actually uses. They are clinical teaching clocks, not a secret ANCC answer key, and individual patients vary — but evaluating earlier than the clock is a classic trap.

If the item says she has taken sertraline 50 mg for 10 days and still feels sad, the answer is not to declare treatment failure and switch to an MAOI. Continue, assess adherence and suicide risk, and re-evaluate at a fair interval. If the item says day-2 amoxicillin for otitis and the toddler is now lethargic with a bulging fontanelle, you do not wait for the 48–72-hour clock — you have the wrong diagnosis or a complication.

Pregnancy changes clocks. Levothyroxine often needs an early dose increase and TSH about every 4 weeks until stable in pregnancy — do not wait a nonpregnant 8 weeks if you just changed the dose in the first trimester. A postpartum SSRI still uses the 4–6-week efficacy clock, but you evaluate infant sedation and maternal suicide risk now.

Safety labs that are part of the outcome

Outcome is not only the disease number. It is also the tax the drug extracted.

ACE inhibitor / ARB / ARNI / mineralocorticoid-receptor antagonist. Recheck creatinine and potassium in 1–2 weeks after start or up-titration, sooner in CKD, heart failure, or dual blockade. A creatinine rise up to about 30% that then stabilizes can be hemodynamic and acceptable; a larger jump, oliguria, or K+ at or above 5.5 mEq/L is an adverse-effect outcome — hold, find the NSAID or dehydrating illness, and do not add a potassium-sparing agent on top. Chapter 9.2 taught the pair. Evaluation is the lab in front of you.

Statin. Routine serial CK and LFTs in an asymptomatic patient are not the modern default. Symptoms drive the workup: new proximal muscle pain or weakness → hold the statin, check CK, look for interacting drugs (strong CYP3A4 inhibitors with simvastatin or lovastatin). Transaminase elevation more than 3× ULN with symptoms, or jaundice, is a stop, not a pep talk. An asymptomatic ALT of 1.3× ULN at week 6 is usually continue-and-repeat, not a lifelong statin ban.

Metformin. Gastrointestinal effects are early and often dose-related; they are adverse effects, not “no glycemic effect.” The late evaluation outcome is vitamin B12 deficiency with long-term use — check B12 if anemia, neuropathy, or unexplained fatigue appears, and periodically in long-term users. Lactic acidosis is rare; the evaluation trigger is tissue hypoxia, severe CKD, or acute kidney injury, not a slightly low bicarbonate on a well patient. Do not start metformin if eGFR is under 30; that is Implementation. Evaluation is the person whose eGFR just fell through 30 on a drug they have taken for years — stop it.

Levothyroxine. Outcome is TSH in range for that person (tighter and trimester-specific in pregnancy; often a bit higher as a target in older adults). Over-replacement in a frail patient with atrial fibrillation is an adverse effect, not “better numbers.”

Warfarin. Therapeutic effect is the INR in the indicated range (typically 2–3 for most indications; 2.5–3.5 for some mechanical valves — know the indication). An INR of 1.4 is no effect (or nonadherence, or a new inducer). An INR of 8 with melena is an adverse effect. Do not congratulate a “therapeutic” INR if the patient is bleeding. Many FNPs now see more DOACs than warfarin; if warfarin is still on the list, the INR remains the outcome lab.

Lithium, if present on the FNP’s list: trough level, TSH, creatinine/eGFR, and pregnancy status are outcome labs, not optional extras. A “fine” mood with a TSH of 18 is not a complete evaluation.

Adherence is the first “failure” explanation

Before you intensify or switch, ask whether the drug was taken. The cheapest reasons a BP is still 162/98 or an A1c is still 9.4 are: never filled, filled once, took it for a week, took it every other day because of cost, took it with a binding interaction (cholestyramine, calcium, food for levothyroxine), or stopped because of an unexplained side effect they were afraid to mention.

Evaluate adherence with nonjudgmental, specific questions, refill histories, and pill counts when needed — not “you’re not trying.” Cost, once-daily versus twice-daily, a 90-day mail-order fill, a combination pill, and a language-concordant handout are evaluation interventions. Health literacy is an outcome modifier: if the patient thought “take two tablets daily” meant two tablets on the day they remembered, the drug did not fail.

White-coat effect, wrong cuff size, and home-BP technique can fake a “failed” antihypertensive. A1c can be misleading in anemia, recent transfusion, and some hemoglobinopathies — pair it with home glucose or CGM before you stack a fourth agent. A parent who shares an albuterol inhaler among three children will look like “uncontrolled asthma on therapy.”

Intensify versus switch versus stop

Use the bucket and the clock.

Intensify (higher dose or add-on from the same strategy) when there is partial therapeutic effect, no dose-limiting adverse effect, and room left in the evidence-based dose range. Example: amlodipine 5 mg, home-average BP 148/90 at 4 weeks, no edema → 10 mg or add a thiazide, depending on the compelling indication.

Switch when there is no meaningful effect at a fair dose and interval, or when a class-specific adverse effect blocks titration (ACE-inhibitor cough → ARB; statin myalgia on simvastatin → a different statin at a proven dose, or a nonstatin if they cannot tolerate any). Switching because the patient “wants something stronger” at day 5 is not evaluation. Switching an SSRI at week 2 for residual sadness, without a suicide or mania signal, is the same error.

Stop when harm exceeds benefit, when the indication has resolved, when pregnancy or a new diagnosis makes the drug contraindicated, or when you have completed a planned finite course (antibiotics, a steroid burst, PPIs that were never deprescribed). Stopping a drug that is working because a Beers flag appeared, without an alternative, is not automatically safer (Chapter 9.2). Stopping metformin for a single GI complaint without a trial of extended-release or with-food dosing is often premature.

Pediatric and older-adult twists

In a preschooler, “no effect” of an antibiotic at 48 hours plus new lethargy is emergency re-evaluation, not a second prescription called to the pharmacy. In an adolescent on an SSRI, the early outcome you cannot skip is suicidality, even if the PHQ-9 has not had time to move. In a frail 90-year-old, a “therapeutic” systolic BP of 108 with near-falls is an adverse effect; loosen the regimen (Chapter 20.2 will take deprescribing further).

Antibiotic “failure” is three different problems

At 48–72 hours of appropriate therapy, if the patient is not improving, do not blindly extend the same prescription. Sort:

1. Wrong diagnosis — viral syndrome, heart failure called “pneumonia,” gout called cellulitis, pulmonary embolism called bronchitis, retained foreign body called “recurrent cellulitis.”
2. Source not drained — abscess, empyema, obstructed kidney, septic joint. Antibiotics do not replace incision and drainage (Chapter 17.1).
3. Resistance or inadequate regimen — wrong drug, wrong dose, poor absorption, or true resistance. Culture if you still can; change therapy with a reason.

Red-flag physiology (hypotension, new confusion, rapidly spreading erythema, orbital findings, a stiff neck after otitis) is not a “give it another day” evaluation. That is emergency transfer. Rechecking a well-appearing adult with cystitis who is dysuria-free at 48 hours is a therapeutic effect — do not extend fluoroquinolones “just in case.”

If you remember one pharmacotherapeutic-outcome sentence: name the bucket, honor the clock, blame adherence before the molecule, and drain what drugs cannot reach.