Yoong Sze Lin, Carey Mariko Leanne, Sanson-Fisher Robert William, D'Este Catherine Anne, Mackenzie Lisa, Boyes Allison
Priority Research Centre for Health Behaviour and Hunter Medical Research Institute, The University of Newcastle, Callaghan, NSW, 2308, Australia,
J Gen Intern Med. 2014 Feb;29(2):328-34. doi: 10.1007/s11606-013-2637-4. Epub 2013 Oct 8.
Overweight and obese patients attempt weight loss when advised to do so by their physicians; however, only a small proportion of these patients report receiving such advice. One reason may be that physicians do not identify their overweight and obese patients.
We aimed to determine the extent that Australian general practitioners (GP) recognise overweight or obesity in their patients, and to explore patient and GP characteristics associated with non-detection of overweight and obesity.
Consenting adult patients (n = 1,111) reported weight, height, demographics and health conditions using a touchscreen computer. GPs (n = 51) completed hard-copy questionnaires indicating whether their patients were overweight or obese. We calculated the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for GP detection, using patient self-reported weight and height as the criterion measure for overweight and obesity. For a subsample of patients (n = 107), we did a sensitivity analysis with patient-measured weight and height. We conducted an adjusted, multivariable logistic regression to explore characteristics associated with non-detection, using random effects to adjust for correlation within GPs.
Sensitivity for GP assessment was 63 % [95 % CI 57-69 %], specificity 89 % [95 % CI 85-92 %], PPV 87 % [95 % CI 83-90 %] and NPV 69 % [95 % CI 65-72 %]. Sensitivity increased by 3 % and specificity was unchanged in the sensitivity analysis. Men (OR: 1.7 [95 % CI 1.1-2.7]), patients without high blood pressure (OR: 1.8 [95 % CI 1.2-2.8]) and without type 2 diabetes (OR: 2.4 [95 % CI 1.2-8.0]) had higher odds of non-detection. Individuals with obesity (OR: 0.1 [95 % CI 0.07-0.2]) or diploma-level education (OR: 0.3 [95%CI 0.1-0.6]) had lower odds of not being identified. No GP characteristics were associated with non-detection of overweight or obesity.
GPs missed identifying a substantial proportion of overweight and obese patients. Strategies to support GPs in identifying their overweight or obese patients need to be implemented.
超重和肥胖患者在医生建议下会尝试减肥；然而，只有一小部分此类患者报告曾收到过此类建议。一个原因可能是医生未能识别出他们的超重和肥胖患者。
我们旨在确定澳大利亚全科医生（GP）识别患者超重或肥胖的程度，并探讨与未检测出超重和肥胖相关的患者及全科医生特征。
同意参与的成年患者（n = 1111）使用触摸屏电脑报告体重、身高、人口统计学信息和健康状况。全科医生（n = 51）填写纸质问卷，表明其患者是否超重或肥胖。我们以患者自我报告的体重和身高作为超重和肥胖的标准测量值，计算全科医生检测的灵敏度、特异度、阳性预测值（PPV）和阴性预测值（NPV）。对于部分患者子样本（n = 107），我们使用患者测量的体重和身高进行了灵敏度分析。我们进行了调整后的多变量逻辑回归，以探讨与未检测出相关的特征，并使用随机效应来调整全科医生内部的相关性。
全科医生评估的灵敏度为63%[95%置信区间57 - 69%]，特异度为89%[95%置信区间85 - 92%]，阳性预测值为87%[95%置信区间83 - 90%]，阴性预测值为69%[95%置信区间65 - 72%]。在灵敏度分析中，灵敏度提高了3%，特异度未变。男性（比值比：1.7[95%置信区间1.1 - 2.7]）、无高血压患者（比值比：1.8[95%置信区间1.2 - 2.8]）和无2型糖尿病患者（比值比：2.4[95%置信区间1.2 - 8.0]）未被检测出的几率更高。肥胖个体（比值比：0.1[95%置信区间0.
