NRNP 6539 Clinical Practice Guidelines Assignment: Evaluating an AI-Generated Hypertension Plan Against the 2025 AHA/ACC Guideline

NRNP 6539: Evaluating an AI-generated clinical case against a clinical practice guideline (assignment template)

• Introduction: in 2–3 sentences, explain that the paper evaluates an AI-generated clinical case and management plan by comparing it with an evidence-based clinical practice guideline, and why applying guidelines matters when evaluating AI-supported decisions.
• Summary of the AI-generated case and recommendations: demographics, chief complaint, history, exam, diagnostic considerations and the AI’s diagnostics, medications, referrals and follow-up (1–2 paragraphs).
• Clinical practice guideline used: name, authoring body, year, link or DOI and 3–4 sentences on why it is appropriate.
• Comparison: where the AI aligns with, differs from or omits guideline recommendations.
• Evaluation of clinical accuracy and safety: appropriateness, safety concerns, diagnostics, medications and referrals, omissions.
• Reflection: AI’s role in advanced practice nursing, risks and responsible use.
• Conclusion, references and the AI Assessment Appendix (prompts, outputs, tool, date, how outputs were checked and used).

Original example for study purposes (about 1,100 words), using hypertension and the 2025 multisociety high blood pressure guideline.

Introduction

This paper evaluates an AI-generated primary care case and management plan for a patient with newly identified hypertension by comparing it with the 2025 AHA/ACC multisociety guideline for high blood pressure in adults. Artificial intelligence tools can produce plausible-sounding clinical plans quickly, but they may rely on outdated or incomplete information. Applying current evidence-based guidelines allows nurse practitioners to verify AI recommendations and protect patient safety.

Summary of the AI-Generated Clinical Case and Recommendations

Case (illustrative example). The AI tool generated the case of a 54-year-old Black man presenting for an annual physical. He reports occasional headaches and no chest pain or dyspnea. History includes obesity (BMI 33), prediabetes (A1c 6.0%) and a father with a stroke at age 62. He drinks three to four beers on weekends and eats fast food several times per week. Office blood pressure averaged 152/96 mm Hg on two readings at this visit, and he reported a similar reading at a pharmacy kiosk last month. Examination was otherwise unremarkable, with no signs of heart failure or retinopathy. The AI listed primary (essential) hypertension as the leading diagnosis and considered white-coat hypertension and secondary causes.

AI recommendations. The AI diagnosed stage 2 hypertension and recommended starting lisinopril 10 mg daily, lifestyle changes (sodium below 2,300 mg/day, weight loss and exercise), a basic metabolic panel and lipid panel, a return visit in three months and a blood pressure goal of less than 140/90 mm Hg. It did not mention home blood pressure monitoring, urine albumin testing, an ECG or cardiovascular risk calculation.

Clinical Practice Guideline Used for Comparison

The comparison uses the 2025 AHA/ACC/AANP/AAPA/ABC/ACCP/ACPM/AGS/AMA/ASPC/NMA/PCNA/SGIM Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults, developed by the American Heart Association, the American College of Cardiology and 11 partner organizations, including the American Association of Nurse Practitioners (Jones et al., 2025; https://doi.org/10.1016/j.jacc.2025.05.007). This guideline is appropriate because it is the most current U.S. evidence-based standard for diagnosing and managing hypertension in adults, it replaced the 2017 guideline and it was developed with nurse practitioner representation. It addresses blood pressure classification, risk assessment, laboratory evaluation, lifestyle therapy, medication selection, treatment targets and follow-up, all of which apply to this primary care case.

Comparison of AI Recommendations to the Guideline

Areas of alignment. The AI correctly classified a blood pressure of 152/96 mm Hg as stage 2 hypertension, which the guideline defines as a systolic pressure of 140 mm Hg or higher or a diastolic pressure of 90 mm Hg or higher. It appropriately recommended lifestyle changes and medication, since the guideline recommends medication plus lifestyle changes for adults averaging 140/90 mm Hg or higher (American Heart Association [AHA], 2025a). An ACE inhibitor is one of the first-line drug classes, along with ARBs, long-acting dihydropyridine calcium channel blockers and thiazide-type diuretics.

Areas of difference. Three important differences emerged. First, the guideline recommends starting stage 2 hypertension with two first-line agents from different classes, preferably as a single-pill combination, rather than monotherapy (AHA, 2025a). Second, the AI’s target of less than 140/90 mm Hg is outdated; the 2025 guideline recommends a target of less than 130/80 mm Hg for most adults. Third, the guideline encourages limiting sodium to less than 2,300 mg/day with an ideal of 1,500 mg/day and recommends ideally avoiding alcohol, or limiting it to two drinks per day for men (AHA, 2025b), while the AI addressed sodium only partially and ignored alcohol.

Missing or incomplete recommendations. The AI omitted confirmation of the diagnosis with out-of-office measurements and home blood pressure monitoring paired with team follow-up; the urine albumin-to-creatinine ratio, which is now recommended for all patients with hypertension; an ECG and other baseline laboratory tests; and cardiovascular risk assessment using the PREVENT equations. It also did not address screening for secondary causes such as primary aldosteronism, for which the guideline expands testing, including in patients with obstructive sleep apnea, or screening for sleep apnea in a patient with obesity. Finally, a three-month follow-up interval is too long for a patient starting treatment for stage 2 hypertension; frequent follow-up and medication titration are needed until the target is reached.

Evaluation of Clinical Accuracy and Safety

The AI’s plan was partially appropriate: it identified the diagnosis and initiated treatment with a safe first-line medication. However, several features reduce its clinical quality and raise safety concerns. Monotherapy and a higher target would likely leave this patient’s blood pressure uncontrolled for months, increasing his risk of stroke, heart failure and kidney disease, particularly given his family history of early stroke. Starting an ACE inhibitor requires baseline and follow-up potassium and creatinine checks, which the AI only partly addressed, and the patient should be counseled about angioedema, which occurs more often in Black patients taking ACE inhibitors. Earlier U.S. guidance favored a thiazide-type diuretic or calcium channel blocker as part of initial therapy for Black adults without compelling indications; the provider should confirm current recommendations for this population in the 2025 guideline. A combination such as an ACE inhibitor or ARB with a calcium channel blocker or thiazide would address both concerns. The omission of home monitoring and risk calculation means the AI could not confirm the diagnosis, rule out white-coat hypertension or guide shared decision-making. Overall, the AI produced a reasonable starting point but not a guideline-concordant plan.

Reflection

AI can support advanced practice nurses by quickly generating differential diagnoses, summarizing guidelines, drafting patient education and flagging drug interactions, which may save time in busy primary care settings. However, AI tools can produce outdated or fabricated recommendations, sound more certain than the evidence warrants and fail to account for individual patient factors, social determinants or patient preferences. They may also reflect biases in their training data. Nurse practitioners can use AI responsibly by treating its outputs as suggestions to verify, checking recommendations against current guidelines and primary evidence, protecting patient privacy by never entering identifiable information into unapproved tools, documenting how AI was used and keeping clinical judgment and accountability with the provider. As this case shows, an AI plan can be mostly right yet miss changes in the latest guideline that matter for patient outcomes.

Conclusion

The AI-generated plan correctly identified stage 2 hypertension and recommended lifestyle changes and a first-line medication, but it used an outdated blood pressure target, recommended monotherapy instead of a single-pill combination, set an overly long follow-up interval and omitted home monitoring, urine albumin testing, ECG and risk assessment. Comparing AI recommendations with current evidence-based guidelines is essential to ensure safe, high-quality care.

References

American Heart Association. (2025a, August 14). Top things to know: 2025 high blood pressure guideline. https://professional.heart.org/en/science-news/2025-high-blood-pressure-guideline/top-things-to-know

American Heart Association. (2025b, August 14). New high blood pressure guideline emphasizes prevention, early treatment to reduce CVD risk [News release]. EurekAlert! https://www.eurekalert.org/news-releases/1094543

Jones, D. W., Ferdinand, K. C., Taler, S. J., Johnson, H. M., Shimbo, D., Abdalla, M., Altieri, M. M., Bansal, N., Bello, N. A., Bress, A. P., Carter, J., Cohen, J. B., Collins, K. J., Commodore-Mensah, Y., Davis, L. L., Egan, B., Khan, S. S., Lloyd-Jones, D. M., Melnyk, B. M., . . . Ziai, W. C. (2025). 2025 AHA/ACC/AANP/AAPA/ABC/ACCP/ACPM/AGS/AMA/ASPC/NMA/PCNA/SGIM guideline for the prevention, detection, evaluation, and management of high blood pressure in adults. Journal of the American College of Cardiology, 86(18), 1567–1678. https://doi.org/10.1016/j.jacc.2025.05.007

More model assignments posted today

• CJUS 801 evaluability assessment (drug court)
• Serial killer analysis: Jeffrey Dahmer
• Profiling an alleyway homicide
• Avery's Gym ERD and normalization