Recurrent abdominal pain that responds to a PPI isn't always a gastric problem

A patient with recurrent upper abdominal pain feels better on a proton pump inhibitor, and the chart quietly closes as "GERD, responding to treatment." That relief is a treatment outcome, not a diagnosis — and at least six other conditions can produce the same pattern, three of them serious enough to matter.

45–54%

Pooled specificity of a PPI symptom trial against endoscopy/pH-metry-confirmed GERD — meaning roughly half of "responders" don't have objectively confirmed reflux disease.^1,2

~1 in 5

Patients on placebo in PPI-trial studies who also reported meaningful symptom relief — a reminder that improvement alone doesn't prove an acid-related mechanism.³

Sensitivity of the PPI trial is reasonably good (~71–79%) — a non-response is informative. A response is not.^1,2

Six origins, one presentation

Epigastric pain that eases on a PPI can come from true reflux — or from anywhere else on this map. The diagnosis is made by pattern-matching the history and exam, not by the drug's effect.

Recurrent epigastric pain — eases on a PPI —

GERD, confirmedErosive change or +pH study

Functional dyspepsiaNormal scope, Rome IV pattern

Biliary colicFatty-meal trigger, RUQ/back

Abdominal wall (ACNES)Point-tender, +Carnett's sign

Atypical angina / ACSExertional; diabetic, elderly, female

Gastric/esophageal cancerAge 40+, new-onset, alarm signs

Confirmed acid-related Functional overlap Mechanical / structural Must actively exclude

If it isn't GERD, here's what tends to hide behind it

Condition	What points away from GERD	What to do next
Functional dyspepsia overlap	Early satiety, postprandial fullness; endoscopy is normal; partial, inconsistent PPI response.	Add/trial a prokinetic; avoid escalating PPI dose indefinitely.
Biliary colic mechanical	Episodic, 30 min–6 h, follows a fatty meal, radiates to back/right scapula.	Abdominal USG — even if the PPI seems to be helping.
Chronic pancreatitis mechanical	Boring pain radiating to the back, worse supine, ± weight loss/steatorrhoea.	Serum lipase; imaging if pain persists beyond one PPI cycle.
Abdominal wall pain (ACNES) mechanical	Sharp, reproducible, one-finger point tenderness; unrelated to meals; positive Carnett's sign.	Bedside Carnett's test (see below); local anaesthetic block if positive.
Atypical angina / ACS exclude	"Indigestion" that's exertional; more common in diabetics, the elderly, and women.⁷	ECG before extending an empirical PPI trial in at-risk patients.
Gastric / esophageal malignancy exclude	Age 40+, new-onset symptoms, any alarm feature — acid suppression can blunt an ulcer-related component of tumour pain while the lesion progresses.	Endoscopy referral — don't let a good response postpone it.

Refer for endoscopy regardless of PPI response, if:

New-onset symptoms at age 40–45+^4,5
Unintentional weight loss
GI bleeding, or unexplained anaemia
Progressive dysphagia / odynophagia
Persistent vomiting
Palpable mass or lymphadenopathy
Family history of upper-GI malignancy

Western guidelines set the age threshold at 45–60; a North Indian OPD dyspepsia series found gastric cancer in 6.7% of endoscoped patients — all aged 40 or above — a reasonable case for a lower local threshold.⁶

A 30-second bedside test worth doing today

Carnett's sign: palpate the tender point, then ask the patient to tense the abdominal wall (lift the head, or both legs, off the table) while you keep palpating. Pain that worsens or persists = abdominal wall source (think ACNES). Pain that eases = likely visceral. No equipment, and it reclassifies patients who've been cycled through years of antacid therapy for a "functional" label.^8,9

Take-home

A positive PPI trial confirms symptom control, not gastric origin — pooled specificity for true GERD is only ~45–54%.
New-onset pain at 40+, or any alarm feature, earns an endoscopy referral alongside — not after — empirical therapy.
A 4–8 week empirical trial that quietly becomes a 4–8 month refill habit is where the misses happen. Re-examine before you renew.
History (meal timing, fatty-food trigger, exertional component, positional change) plus Carnett's sign filters out half the mimics without an investigation.

Healthcity Hospital keeps outpatient endoscopy, abdominal ultrasound, and GI consultation slots available for fast-tracked workup on patients you flag from this list. Reach out through the Med-Connect group or your usual coordinator to arrange a slot.

References

Numans ME, Lau J, de Wit NJ, Bonis PA. Short-term treatment with proton-pump inhibitors as a test for GERD: a meta-analysis of diagnostic test characteristics. Ann Intern Med. 2004;140(7):518–27.
Ghoneim S, Wang J, El Hage Chehade N, et al. Diagnostic accuracy of the PPI test in GERD and non-cardiac chest pain: a systematic review and meta-analysis. J Clin Gastroenterol. 2023;57(4):380–8.
Is proton pump inhibitor testing an effective approach to diagnose GERD in patients with non-cardiac chest pain: a meta-analysis [structured abstract]. Database of Abstracts of Reviews of Effects (DARE). York: Centre for Reviews and Dissemination.
American Gastroenterological Association. AGA medical position statement: evaluation of dyspepsia. Gastroenterology. 1998;114(3):579–81.
Moayyedi PM, Lacy BE, Andrews CN, et al. ACG and CAG clinical guideline: management of dyspepsia. Am J Gastroenterol. 2017;112(7):988–1013.
Prevalence of Helicobacter pylori infection in upper GI tract disorders (dyspepsia) patients visiting the OPD of a hospital in North India. J Family Med Prim Care. 2018. doi:10.4103/jfmpc.jfmpc_213_17.
Typical and atypical symptoms of acute coronary syndrome: time to retire the terms? J Am Heart Assoc. 2020;9(1):e015539.
Van Assen T, de Jager-Kievit JWAM, Scheltinga MRM, Roumen RMH. Chronic abdominal wall pain misdiagnosed as functional abdominal pain. J Am Board Fam Med. 2013;26(6):738–44.
Scheltinga MR, Roumen RM. Anterior cutaneous nerve entrapment syndrome (ACNES). Hernia. 2018;22(3):507–16.

Healthcity Med-Connect · GI Series, Issue 01 of 6 For clinical education among network physicians

Dr. Yogesh Dukare