Avoiding loss of native individuals in birth certificate data.
Avoiding loss of native individuals in birth certificate data.
复制标题
DOI：
10.1038/s41372-022-01469-4
复制
发表时间：
2023-03
影响因子：
2.9
通讯作者：
Montoya-Williams, Diana
中科院分区：
文献类型：
--
作者：
Holloway, Kayla;Radack, Joshua;Passarella, Molly;Ellison, Angela M.;Chaiyachati, Barbara H.;Burris, Heather H.;Montoya-Williams, Diana
与非西班牙裔白色家庭相比，美洲印第安人或阿拉斯加原住民（AIAN）家庭受到早产的影响尤为严重（2020年分别为11.6%和9.1%）[1]。在整个生命周期中持续存在的不平等导致AIAN个体的预期寿命比美国任何其他种族群体都短[2]。改善AIAN家庭健康和医疗保健的一个挑战是缺乏研究代表性[3]。这部分是由于AIAN个体在美国人口中所占比例很小（约占2%）。虽然分析健康方面种族差异的金标准是使用自我认同的种族[4]，但先进的统计模型通常无法适应非常小的样本量[5]。因此，AIAN个体在分析包括早产在内的健康结果时经常被吸收到“其他”类别中。这种汇总可能掩盖可采取行动的差异。在宾夕法尼亚州（PA）的出生证明上，父母可以通过两种方式表明他们的种族和民族。父母可以勾选复选框，他们可以使用“写入”选项。本研究的目的是确定是否使用写入数据可以有意义地增加围产期研究中AIAN人群的样本量。我们使用了2006年至2014年的PA出生证明数据，仅限于与出院数据相关的单胎出生，以确定通过检查每个写入响应，我们可以捕获AIAN中有多少额外的出生。然后，我们研究了早产率及其95%置信区间在AIAN人群中的差异，当限制在选中AIAN框的个体或扩展到包括写入响应时。最后，我们进行了把握度计算，以证明增加样本量对研究相对罕见的事件（如早产）的重要性。在研究期间有1，089，429名新生儿，其中905名（0.08%）是只选中AIAN框的人。另有11895人（1.1%）在另一个框中选中了AIAN框，他们被归类为多种族，未纳入分析。近一半的这些多种族的个人（n= 5277）确定为黑色除了AIAN。有222人勾选了“其他”框，并写下了与AIAN身份一致的回复。最常见的写入响应按降序排列为：切罗基（n= 50）、泰诺（n= 23）、美洲土著（n= 21）和黑脚（n= 8）。所有其他书面回答的频率均为1。包括这些答复增加了样本量，1109个AIAN个体之间的出生，增加了22.5%。AIAN人群的早产率（不包括填写回复）为10.1%（95% CI：8.1%，12.0%）。当纳入写入回答时，早产率略高，预期置信区间缩小（10.8%，95% CI：9.0%，12.7%）。功效曲线显示，当使用较大的组时，检测早产率相对变化的功效有所提高，α为0.05（图1）。使用这个数字，我们可以看到，如果研究人员希望测试，例如，早产率是否增加了20%，那么1109名新生儿的较大组将比905名新生儿的原始组具有大约10%的能力来检测这种增加。总之，在分析出生证数据时列入土著印第安人和土著民族的书面指定，有很大的潜力增加代表性，避免信息丢失。此外，未来将AIAN纳入其身份的个人可能会揭示健康受结构性疾病影响的交叉方式。
American Indian or Alaskan Native (AIAN) families are disproportionately affected by preterm birth compared to non-Hispanic White families (11.6 vs. 9.1% in 2020)[1]. Persistent inequities across the lifespan result in shorter life expectancies among AIAN individuals than any other racial group in the United States [2]. One challenge to improving health and healthcare of AIAN families is the lack of representation in research [3]. This is partly due to the small proportion of AIAN individuals in the US population (constituting roughly two percent). While the gold standard in analyses of racial disparities in health is to use selfidentified race [4], advanced statistical models often cannot accommodate very small sample sizes [5]. As such, AIAN individuals often are absorbed into an “other” category when analyzing health outcomes, including preterm birth. Such aggregation can mask actionable disparities. On the Pennsylvania (PA) birth certificate, there are two ways in which parents can indicate their race and ethnicity. Parents can check boxes and they can use a “write-in” option. The aim of this study was to determine whether using write-in data could meaningfully increase the sample size of AIAN populations in perinatal research. We used PA birth certificate data from 2006 to 2014, restricted to singleton births linked to hospital discharge data, to determine how many additional births among AIAN we could capture with examination of each write-in response. We then examined how preterm birth rates and their 95% confidence intervals differed among AIAN people when restricting to individuals who checked the AIAN box or when expanding to include write-in responses. Finally, we performed a power calculation to demonstrate the importance of increasing sample sizes to study relatively rare events such as preterm birth. There were 1,089,429 births in the study period, 905 (0.08%) of which were among individuals who checked only the AIAN box. There were an additional 11 895 (1.1%) individuals who checked the AIAN box in addition to another box, who were classified as multiracial and not included in the analysis. Nearly half of these multiracial individuals (n= 5277) identified as Black in addition to AIAN. There were 222 individuals who checked the “other” box and wrote-in a response consistent with AIAN identities. The most common write-in responses in descending order were: Cherokee (n= 50), Taino (n= 23), Native American (n= 21), and Blackfoot (n= 8). All other written-in responses had a frequency of one. Including these responses increased the sample size to 1109 births among AIAN individuals–a 22.5% increase. The preterm birth rate among AIAN people without including write-in responses was 10.1%(95% CI: 8.1%, 12.0%). When write-in responses were included, the preterm birth rate was slightly higher with an expected narrowing of the confidence interval (10.8%, 95% CI: 9.0%, 12.7%). A power curve shows the improvement in power to detect relative changes in preterm birth rates with an alpha of 0.05 when the larger group is used (Fig. 1). Using this figure, one can see that if researchers were hoping to test, for instance, whether the preterm birth rate had increased by 20%, the larger group of 1109 births would have roughly 10% more power than the original group of 905 births to detect this increase. In conclusion, inclusion of write-in designations of AIAN populations in analyses of birth certificate data has substantial potential to increase representation and avoid loss of information. In addition, future inclusion of individuals who include AIAN among their identities may shed light on intersectional ways in which health is affected by structural …
影响因子：
12.7
作者：
Pacheco CM;Daley SM;Brown T;Filippi M;Greiner KA;Daley CM
通讯作者：
Daley CM
影响因子：
4.5
作者：
Moradi, Milad;Samwald, Matthias
通讯作者：
Samwald, Matthias