Table 4. Domain-specific GIRISH prompt templates for neonatal clinical practice.
AAP: American Academy of Pediatrics; aEEG: Amplitude-integrated electroencephalography; AUC: Area under the curve; AXR: Abdominal radiograph; BP: Blood pressure; BPD: Bronchopulmonary dysplasia; BW: Birth weight; CHD: Congenital heart disease; CMV: Cytomegalovirus; CRP: C-reactive protein; CVL: Central venous line; CXR: Chest radiograph; DHA: Docosahexaenoic acid; DIC: Disseminated intravascular coagulation; DOL: Day of life; EBM: Expressed breast milk; ELBW: Extremely low birth weight; EOS: Early-onset sepsis; ESPGHAN: European Society for Paediatric Gastroenterology Hepatology and Nutrition; ETTNO: Effects of transfusion thresholds on neurocognitive outcome trial; FiO2: Fraction of inspired oxygen; GA: Gestational age; GI: Gastrointestinal; GIR: Glucose infusion rate; GORD: Gastro-esophageal reflux disease; HFOV: High-frequency oscillatory ventilation; HIE: Hypoxic-ischemic encephalopathy; HR: Heart rate; IEM: Inborn errors of metabolism; IM: Intramuscular; iNO: Inhaled nitric oxide; IV: Intravenous; IVH: Intraventricular hemorrhage; LA:Ao: Left-atrium-to-aorta ratio; LMIC: Low- and middle-income country; LOS: Late-onset sepsis; MAP: Mean arterial pressure; MRI: Magnetic resonance imaging; NAS: Neonatal abstinence syndrome; NEC: Necrotizing enterocolitis; NICHD: National Institute of Child Health and Human Development; NNF: National Neonatology Forum (India); OI: Oxygenation index; pCO2: Partial pressure of carbon dioxide; PDA: Patent ductus arteriosus; PICC: Peripherally inserted central catheter; PMA: Post-menstrual age; PPHN: Persistent pulmonary hypertension of the newborn; RDS: Respiratory distress syndrome; SIP: Spontaneous intestinal perforation; TOBY: Total body hypothermia for neonatal encephalopathy trial; TOP: Transfusion of prematures trial; TORCH: Toxoplasmosis, other, rubella, cytomegalovirus, herpes simplex; TPN: Total parenteral nutrition; UAC: Umbilical arterial catheter.
| Clinical domain | Goal statement (G) – copy and adapt | Key input parameters (I) | Safety verification (S) – always do this |
| Respiratory (RDS, BPD, PPHN, apnea) | “Summarize evidence for [surfactant/iNO/HFOV vs SIMV] in a [GA]-week infant with [diagnosis]. State evidence grade. Do not prescribe.” | GA, BW, DOL, FiO2, MAP, OI, blood gas (pH, pCO2, base excess), surfactant history, CXR result, ventilator mode and settings, caffeine dose | iNO: OI at or above 25, term/near-term (AAP 2019). Surfactant: weight/GA-specific dosing. HFOV: set MAP, amplitude, frequency. PPHN: exclude sepsis and metabolic causes first. Use WHO guidance for LMIC settings. |
| Cardiovascular (PDA, CHD, shock) | “Evidence for and against medical PDA closure vs expectant management in a [GA]-week infant, [DOL], with the echo findings below. Include evidence grade. Do not prescribe.” | Echo findings (diameter, flow, LA:Ao ratio), HR, BP (MAP), capillary refill, urine output, 24 h fluid balance, respiratory support, prior indomethacin/ibuprofen | Indomethacin: contraindicated if urine output below 1 mL/kg/h. Ibuprofen: contraindicated with active GI bleed. IV paracetamol: dose by GA. Ligation: per unit protocol. Verify with senior or cardiologist. |
| Neurological (HIE, IVH, seizures, NAS) | “Assess therapeutic-hypothermia eligibility (TOBY [23]/NICHD [24] criteria). List criteria met, not met, and uncertainties. State what additional information is needed.” | Mode of birth, time to first cry, Apgar scores (1, 5, 10 min), cord/first-gas pH and base excess, aEEG description, neurological examination, age in hours, GA (must be at or above 36w) | Cooling window: start within 6 h of age. Eligibility: pH at or below 7.00 or base deficit at or above 16 mmol/L, and/or abnormal neurology and/or abnormal aEEG. Phenobarbitone loading: dose by weight, not GA. Levetiracetam: evidence grade B only. MRI: days 3 to 5 optimal. |
| Gastrointestinal (NEC, SIP, GORD, jaundice) | “Stage this infant using modified Bell’s criteria. Recommend initial management. State surgical-referral triggers. Flag evidence grade throughout.” | Abdominal findings (distension, tenderness, erythema, palpable loops), AXR result (pneumatosis, portal gas, free air), WBC, CRP, platelet count and trend, feed history and volume at presentation, stool pattern | Triple antibiotics: ampicillin, gentamicin, and metronidazole (confirm local protocol). Nil by mouth: minimum 7 to 10 days for Bell IIa/IIb. Bell IIb/III: urgent surgical consult. Antifungal prophylaxis per unit protocol in ELBW/less-than-28-week infants in high-prevalence units. Peritoneal drain vs laparotomy: discuss with surgery per local protocol. |
| Infectious disease (EOS, LOS, fungal, TORCH) | “Recommend empirical antibiotics for suspected LOS in the infant below. Include evidence grade and duration. State what requires culture results to determine.” | GA, PMA, DOL, prior antibiotics (agent, duration), culture results, CRP trend (last 3), access device (UAC/PICC/CVL), local antibiogram, signs of a specific focus | Gentamicin: extended-interval dosing by GA (Neofax). Vancomycin: AUC-guided monitoring preferred over trough. Amphotericin B: check renal function first. De-escalate on sensitivities within 48 to 72 hours – prompt the AI to state this. |
| Metabolic (hypoglycemia, IEM, TPN, electrolytes) | “Stepwise investigation algorithm for persistent hypoglycemia unresponsive to GIR above 8 mg/kg/min. Flag diagnoses requiring urgent exclusion.” | Blood-glucose trend (last 6 readings with GIR at each), current GIR, insulin level if measured, cortisol, growth hormone, lactate, ammonia, urine ketones, amino-acid screen, family history, maternal medications, gestational complications | Glucagon 200 mcg/kg IV/IM; blunted response in preterm. Diazoxide: monitor for PPHN. Octreotide: case-series only – specialist input. IEM screen: critical sample during hypoglycemia. Metabolic referral if undiagnosed at 48 hours. |
| Hematologic (anemia, DIC, polycythemia) | “Summarize transfusion-threshold evidence for a [GA/PMA]-week infant on [ventilatory support]. Reference TOP [25] and ETTNO [26] trial results.” | Hemoglobin trend (last 3), respiratory support and FiO2, 24 h apnea frequency, weight trend, reticulocyte count, blood group and Coombs, bilirubin, erythropoietin status | TOP [25] and ETTNO [26] (2020–2022): liberal and restrictive thresholds showed equivalent neurodevelopmental outcomes, with fewer transfusions under the restrictive approach. Use CMV-negative/irradiated products for immunocompromised infants. Platelet threshold: 25 x109/L if stable; 50 x109/L pre-procedure; 100 x109/L with active bleeding. |
| Nutrition and prematurity (TPN, enteral feeding, growth) | “Evidence-based enteral-advancement protocol for a [GA]-week infant, [DOL], with the feeding history below. State feed-intolerance triggers.” | GA, BW, DOL, feed volume (mL/kg/day), feed type (EBM/donor/formula), tolerance indicators, 7-day weight trend, TPN composition, NEC risk factors, fortification status | Advancement rate: up to 30 mL/kg/day increase in the very preterm. Fortification: commence at 100 mL/kg/day. Vitamin D: 400 to 800 IU/day. DHA: evidence exists, but quality is variable. Probiotics: check the national-body position (NNF India, ESPGHAN). |