Background Integration of depression treatment into primary care could improve patient outcomes in low-resource settings. Losses along the depression care cascade limit integrated service effectiveness. This study identified patient-level factors that predicted detection of depressive symptoms by nurses, referral for depression treatment, and uptake of counseling, as part of integrated care in KwaZulu-Natal, South Africa. Methods This was an analysis of baseline data from a prospective cohort. Participants were adult patients with at least moderate depressive symptoms at primary care facilities in Amajuba, KwaZulu-Natal, South Africa. Participants were screened for depressive symptoms prior to routine assessment by a nurse. Generalized linear mixed-effects models were used to estimate associations between patient characteristics and service delivery outcomes. Results Data from 412 participants were analyzed. Nurses successfully detected depressive symptoms in 208 [50.5%, 95% confidence interval (CI) 38.9–62.0] participants; of these, they referred 76 (36.5%, 95% CI 20.3–56.5) for depression treatment; of these, 18 (23.7%, 95% CI 10.7–44.6) attended at least one session of depression counseling. Depressive symptom severity, alcohol use severity, and perceived stress were associated with detection. Similar factors did not drive referral or counseling uptake. Conclusions Nurses detected patients with depressive symptoms at rates comparable to primary care providers in high-resource settings, though gaps in referral and uptake persist. Nurses were more likely to detect symptoms among patients in more severe mental distress. Implementation strategies for integrated mental health care in low-resource settings should target improved rates of detection, referral, and uptake.

中文翻译：

---

南非夸祖鲁-纳塔尔省慢性护理患者抑郁症状检测、转诊和接受抑郁咨询的患者水平预测因素

背景 将抑郁症治疗纳入初级保健可以改善资源匮乏地区的患者预后。抑郁症护理级联的损失限制了综合服务的有效性。作为南非夸祖鲁-纳塔尔省综合护理的一部分，这项研究确定了预测护士检测抑郁症状、转诊抑郁症治疗和接受咨询的患者水平因素。方法 这是对来自前瞻性队列的基线数据的分析。参与者是在南非夸祖鲁-纳塔尔省 Amajuba 的初级保健机构中至少有中度抑郁症状的成年患者。在护士进行常规评估之前，对参与者进行了抑郁症状筛查。广义线性混合效应模型用于估计患者特征与服务提供结果之间的关联。结果 分析了 412 名参与者的数据。护士成功检测到 208 名 [50.5%, 95% 置信区间 (CI) 38.9–62.0] 参与者的抑郁症状；其中，他们将 76 人 (36.5%, 95% CI 20.3–56.5) 转介给抑郁症治疗；其中，18 人（23.7%，95% CI 10.7–44.6）至少参加了一次抑郁症咨询。抑郁症状严重程度、酒精使用严重程度和感知压力与检测相关。类似的因素并没有推动转诊或咨询的采用。结论 护士在资源丰富的地区以与初级保健提供者相当的比率检测出患有抑郁症状的患者，尽管在转诊和吸收方面的差距仍然存在。护士更有可能在精神压力更严重的患者中发现症状。在资源匮乏地区实施综合精神卫生保健的战略应以提高检出率、转诊率和接受率为目标。