Clinical Documents

To edit the draft, click on the text box below.

×

Draft a comprehensive differential diagnosis for acute substernal chest pain in a 45-year-old male with hypertension and hyperlipidemia, diaphoresis, exertional onset, and normal initial ECG/troponin; include life-threatening causes first, key distinguishing features, and immediate diagnostic next steps.

Draft a detailed differential diagnosis for new-onset unilateral throbbing headache with visual aura and brief monocular vision loss in a 32-year-old female on oral contraceptives; include red-flag symptoms, stroke/TIA considerations, and recommended initial workup.

Draft an assessment and plan for stage-2 hypertension (BP 162/98) in an obese 52-year-old with suspected obstructive sleep apnea; include ASCVD risk estimate, lifestyle prescriptions, first-line pharmacotherapy with drug/dose/targets, baseline labs/monitoring, and 4–6 week follow-up.

Draft an assessment and plan for type 2 diabetes follow-up with A1c 9.2% on metformin 1,000 mg BID and eGFR 65; compare adding a GLP-1 receptor agonist vs an SGLT2 inhibitor, address weight loss goals, hypoglycemia education, vaccinations, and micro/macrovascular screening.

Draft a history and physical for a patient presenting with chest pain, including HPI using OLDCARTS, pertinent ROS, PMH/meds/allergies, FH/SH, focused cardiac/pulmonary exam, initial labs and imaging orders, prioritized differential, HEART score risk stratification, and early management.

Draft a history and physical for a new patient with progressive exertional dyspnea, orthopnea, and bilateral leg edema; include differential (heart failure, COPD, PE, anemia, renal disease), targeted exam maneuvers, and first-line diagnostics and management.

Draft a progress note for COPD follow-up two weeks after an acute exacerbation treated with prednisone and antibiotics; include symptom trajectory, CAT score (if available), inhaler technique/adherence, oxygen needs, spirometry plan, smoking cessation counseling, and return precautions.

Draft a progress note for post-operative day 2 after total knee arthroplasty; include pain control strategy, DVT prophylaxis, PT/OT milestones, wound status, bowel/urinary function, labs/vitals review, complications screening, and discharge readiness.

Draft a discharge summary for community-acquired pneumonia admission; include admission/discharge dates, presenting symptoms, radiology/labs, pathogen if identified, inpatient therapies and response, discharge medications with durations, pending tests, follow-up appointments, and return precautions.

Draft a discharge summary for myocardial infarction managed with PCI and drug-eluting stent placement; include cath findings, complications, inpatient course, discharge medications (DAPT, statin, beta-blocker, ACEi/ARB), cardiac rehab referral, and follow-up plan.

Draft discharge instructions for a patient hospitalized with acute decompensated heart failure; include daily weights, sodium and fluid restriction amounts, medication list with dosing, signs of worsening heart failure, follow-up appointments, and lab monitoring plan.

Draft discharge instructions for an asthma exacerbation; include controller and reliever regimen, steroid taper schedule, spacer technique, trigger avoidance, a written asthma action plan, return precautions, and follow-up timing.

Draft a patient handout on managing high blood pressure in plain language; explain what BP numbers mean, key lifestyle steps (DASH diet, sodium limits, alcohol moderation), home BP monitoring instructions, medication adherence tips, and when to call the doctor.

Draft a patient handout for type 2 diabetes education in plain language; cover A1c goals, the plate method, physical activity targets, self-monitoring of blood glucose, hypoglycemia signs and the 15-15 rule, foot care basics, and a vaccination checklist.

Draft a summary of the latest CDC and WHO guidelines for managing community-acquired pneumonia in adults, including empiric antibiotic choices, duration of therapy, and follow-up recommendations.

Draft a plain-language overview of the ADA and AACE guidelines for diabetes management, focusing on A1c goals, weight management, and the use of GLP-1 receptor agonists and SGLT2 inhibitors.

Draft a summary of the 2024 USPSTF breast cancer screening recommendations, including updated screening intervals, risk-based approaches, and the role of MRI for high-risk patients.

Draft a concise summary of NCCN guidelines for melanoma management, including staging, surgical margins, adjuvant therapy, and checkpoint inhibitor recommendations.

Draft treatment alternatives for community-acquired MRSA skin infections in patients allergic to sulfa drugs; include oral and IV options, dosing, and duration.

Draft treatment alternatives for patients with type 2 diabetes who are unable to tolerate GLP-1 receptor agonists due to nausea; include medication substitutions and lifestyle recommendations.

Draft a summary of options for treating refractory depression, comparing ketamine infusions, transcranial magnetic stimulation (TMS), and electroconvulsive therapy (ECT).

Draft treatment options for advanced prostate cancer, including hormonal therapy, radiopharmaceuticals, and novel androgen receptor pathway inhibitors.

Draft a summary of recent research on mRNA vaccine technology beyond COVID-19, including applications in oncology and autoimmune disease prevention.

Draft an easy-to-understand overview of long COVID research, summarizing major symptoms, hypothesized mechanisms, and emerging treatment approaches.

Draft a summary of current research on Alzheimer’s disease biomarkers, including amyloid and tau imaging, and discuss how they influence diagnosis and trial design.

Draft an explanation of the interaction between warfarin and antibiotics such as ciprofloxacin or TMP-SMX, including INR effects and management recommendations.

Draft a summary evaluating drug interactions for patients taking Paxlovid and common cardiovascular drugs; include mechanism and recommended dose adjustments.

Draft a summary of the common and serious side effects of semaglutide, including GI intolerance, gallbladder disease, and thyroid-related warnings.

Draft an overview of potential side effects associated with long-term corticosteroid use, including osteoporosis, glucose intolerance, and adrenal suppression.

Draft a summary of essential lab tests to order in suspected acute pancreatitis, including amylase, lipase, liver enzymes, and metabolic panels, with interpretation guidance.

Draft laboratory workup recommendations for suspected hypothyroidism, including TSH, free T4, thyroid antibodies, and when to consider secondary causes.

Draft an explanation of which labs should be monitored in patients receiving methotrexate therapy, including CBC, LFTs, renal function, and frequency of monitoring.

why choose us

Edit Geo

×

300×250 Ad Slot

Research Article: Performance of four large language model chatbots in responding to patient-facing questions about osteoporotic vertebral compression fractures: a cross-sectional paired evaluation

Date Published: 2026-10-16

Abstract:
Large language model (LLM) chatbots are increasingly used by the public to obtain health information, but the safety and quality of their responses to questions about osteoporotic vertebral compression fractures (OVCFs) remain unclear. To compare the safety and patient-facing response quality of four publicly accessible LLM chatbots when responding to standardized questions about osteoporotic vertebral compression fractures. In this cross-sectional paired evaluation, 36 standardized OVCF-related patient-facing questions were submitted to each of four publicly accessible LLM chatbot interfaces—ChatGPT, Gemini, DeepSeek, and Doubao—in an initial response-generation round on June 20, 2026, from Jiangsu, China. Queries were conducted in English in independent single-turn sessions under matched conditions, yielding 144 responses for the primary between-chatbot analysis. A second independent generation round using the same 36 questions and matched query procedures was conducted 7 days later (June 27, 2026), yielding an additional 144 responses for within-model stability analysis. The questions covered symptoms and warning signs, diagnosis and imaging, conservative management, medication safety, rehabilitation, vertebral augmentation, fracture prevention, and special clinical scenarios. Five clinical raters, blinded to chatbot identity and generation round, independently evaluated safety, factual accuracy, empathic communication, treatment-related information quality using DISCERN, patient-information quality using EQIP, transparency using the JAMA benchmark criteria, and overall quality using the Global Quality Score. Readability was assessed automatically using six established indices. Between-chatbot comparisons and test–retest stability analyses used paired methods appropriate for the outcome type. Inter-rater agreement was high, with a Fleiss’ kappa of 0.818 for safety and ICC(2,1) values of 0.816 for accuracy, 0.820 for empathy, 0.820 for DISCERN, 0.820 for EQIP, 0.821 for JAMA, and 0.818 for GQS. Potentially unsafe responses were identified for every chatbot: 3 of 36 ChatGPT responses (8.3%), 5 of 36 DeepSeek responses (13.9%), 6 of 36 Gemini responses (16.7%), and 7 of 36 Doubao responses (19.4%). The overall paired difference in safety was not statistically significant (Cochran's Q?=?3.621, df?=?3, P =?0.305). Factual accuracy differed significantly but only modestly among chatbots ( P =?0.009; Kendall's W?=?0.108). Empathy showed substantially greater variation ( P <?0.001; W?=?0.817), with higher scores for ChatGPT and Gemini than for DeepSeek and Doubao. DISCERN, EQIP, JAMA benchmark, and GQS scores also differed significantly among chatbots (all P <?0.001). Doubao produced responses with the greatest estimated textual complexity. Across the two generation rounds, safety agreement ranged from 88.9% to 97.2% (Gwet's AC1, 0.838–0.966), and median ICC(2,1) values across the 12 non-safety outcomes ranged from 0.781 to 0.904. In this time-stamped, single-location, two-round evaluation, all four chatbot interfaces produced some responses classified as potentially unsafe despite generally high clinician-rated factual accuracy. Differences were greater in empathy, information quality, transparency, overall quality, and readability than in factual accuracy. The observed within-model stability was generally good but did not eliminate variability across generations. These findings are specific to the evaluated access location, language, single-turn conditions, and generation dates and should not be assumed to represent stable performance across settings or future model updates. LLM-generated OVCF information should not be used as a standalone patient education resource; any potential use should involve clinician oversight, safety safeguards, and further validation.

Introduction:
Large language model (LLM) chatbots are increasingly used by the public to obtain health information, but the safety and quality of their responses to questions about osteoporotic vertebral compression fractures (OVCFs) remain unclear.

Read more

300×250 Ad Slot

×
Chat & Ask Questions Career Openings Medical Doctors Check on Drug
Interactions Latest Health News
by WebHealthNetwork Get CME Credits Latest Clinical Trials Latest Research Articles Clinical Documents Update Profile

News

Chat

HCP Tools

Careers

Update Profile