RESUMEN
BACKGROUND AND AIM: Comprehensive reports on the risk factors for bleeding and early death after percutaneous endoscopic gastrostomy (PEG) are limited. In this multicenter study, we retrospectively investigated the risk factors for bleeding and early death after PEG. METHODS: Patients (n = 1234) who underwent PEG between 2015 and 2020 at Osaka Medical and Pharmaceutical University and its affiliated hospitals (11 institutions in total) were evaluated for postoperative bleeding and early death (within 60 days) after PEG according to patient characteristics, construction method, medical history, medications, preoperative hematological findings, and perioperative adverse events. Multivariate logistic regression was performed to identify independent predictors of bleeding and early death after PEG. RESULTS: The risk factors for bleeding after PEG were PEG tube insertion using the modified introducer method (odds ratio [OR], 4.37; P = 0.0003), low platelet count (OR, 0.99; P = 0.014), antiplatelet therapy (OR, 2.11; P = 0.036), and heparinization (OR, 4.50; P = 0.007). Risk factors for early death were low body mass index (BMI) (OR, 0.89; P = 0.015), low serum albumin levels (OR, 0.50; P = 0.035), and comorbidity of active cancer (OR, 4.03; P < 0.0001). There was no significant association between bleeding and early death after PEG. CONCLUSIONS: We identified several risk factors for bleeding and early death after PEG. Risk factors for bleeding were PEG tube insertion using the modified introducer method, low platelet count, antiplatelet therapy, and heparinization. Risk factors for early death were low BMI, low serum albumin levels, and comorbidity of active cancer.
Asunto(s)
Gastrostomía , Mortalidad Prematura , Hemorragia Posoperatoria , Gastrostomía/efectos adversos , Humanos , Neoplasias/epidemiología , Inhibidores de Agregación Plaquetaria/efectos adversos , Hemorragia Posoperatoria/epidemiología , Estudios Retrospectivos , Factores de Riesgo , Albúmina SéricaRESUMEN
BACKGROUND: Gastroesophageal reflux disease (GERD) can be treated using a vonoprazan-first strategy (first-line treatment with vonoprazan), or esomeprazole-first/rabeprazole-first strategies (first-line treatment with proton-pump inhibitors [PPIs], esomeprazole/rabeprazole, followed by a switch to vonoprazan). This cost-utility analysis used long-term simulation modeling to evaluate the cost-effectiveness of a vonoprazan-first strategy compared with the esomeprazole-first and rabeprazole-first strategies. METHODS: A Markov simulation model was developed to evaluate the cost-effectiveness of vonoprazan-first, esomeprazole-first, and rabeprazole-first strategies, comprising healing and maintenance therapies, over 5 years (4-week cycles). Healing therapy began with the administration of a normal dose of drug per real-world practice. If patients were not healed endoscopically, either a longer duration of healing therapy was provided (vonoprazan), the dose was increased (rabeprazole), or patients were switched to vonoprazan (immediately for esomeprazole, and after dose-escalation for rabeprazole, respectively). Healed patients received maintenance (lower/same dose as healing therapy). Recurrence resulted in re-challenge with healing therapy. Transition probabilities were derived from the results of indirect comparisons (network meta-analysis) and costs calculated from the Japanese payer perspective. Outcomes were defined as quality-adjusted life years (QALYs), with utilities based on published values. RESULTS: Expected costs of the vonoprazan-, esomeprazole-, and rabeprazole-first strategies were ¥36,194, ¥76,719, and ¥41,105, respectively, over 5 years. QALY gains for vonoprazan-first strategy versus the esomeprazole- and rabeprazole-first strategies were 0.014 and 0.003, respectively. Both estimated incremental cost-effectiveness ratios were dominant and robust to two sensitivity analyses. CONCLUSIONS: Vonoprazan-first strategy increased QALYs and appeared to be cost-effective for GERD patients compared with the esomeprazole- or rabeprazole-first strategies.