R的流行率的Meta分析
目前，我正在研究流行率的荟萃分析。在R中有很多用于元分析的包（即meta，metafor，MAd，......）。尽管如此，我无法找到适当的流行率指令。您对R中的荟萃分析和患病率有任何经验吗？你会用哪个命令？
library(metafor)
res <- rma(p(logit), SE(logit), data=data, method="DL")
forest(res)
有什么想法吗？
祝你好运 托蒂
2 个答案:
答案 0 :(得分：1)
R包metaprop的命令meta可以执行比例的元分析。它也已在Stata中实施
有关其使用的好教程here。这项工作展示了如何使用metaprop的Stata版本，但R版本的许多内容也是如此
下面我将展示文章中考虑的一个数据集的元分析：子宫颈抹片检查显示ASC-US的女性HPV感染的患病率。
这是数据集：
df <- structure(list(study = c("Manos, 1999", "Bergeron, 2000", "Lytwyn, 2000",
"Shlay, 2000", "Morin, 2001", "Rebello, 2001", "Solomon, 2001",
"Zielinski, 2001", "Kulasingam, 2002", "Pretorius, 2002", "Cuzick, 2003",
"Guyot, 2003", "Lonky, 2003", "Wensveen, 2003", "Bruner, 2004",
"Rowe, 2004", "Andersson, 2005", "Dalla Palma, 2005", "Giovannelli, 2005",
"Kendall, 2005", "Nieh, 2005", "Bergeron, 2006", "Kelly, 2006",
"Kiatpongsan, 2006", "Ko, 2006", "Monsonego, 2006", "Moss, 2006",
"Selvaggi, 2006", "Wright, 2006", "Cuschieri, 2007", "Ronco, 2007",
"You, 2007"), author = c("Manos", "Bergeron", "Lytwyn", "Shlay",
"Morin", "Rebello", "Solomon", "Zielinski", "Kulasingam", "Pretorius",
"Cuzick", "Guyot", "Lonky", "Wensveen", "Bruner", "Rowe", "Andersson",
"Palma", "Giovannelli", "Kendall", "Nieh", "Bergeron", "Kelly",
"Kiatpongsan", "Ko", "Monsonego", "Moss", "Selvaggi", "Wright",
"Cuschieri", "Ronco", "You"), year = c(1999L, 2000L, 2000L, 2000L,
2001L, 2001L, 2001L, 2001L, 2002L, 2002L, 2003L, 2003L, 2003L,
2003L, 2004L, 2004L, 2005L, 2005L, 2005L, 2005L, 2005L, 2006L,
2006L, 2006L, 2006L, 2006L, 2006L, 2006L, 2006L, 2007L, 2007L,
2007L), tgroup = structure(c(1L, 1L, 1L, 1L, 1L, 9L, 1L, 9L,
1L, 1L, 9L, 9L, 1L, 1L, 6L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 6L, 1L,
6L, 1L, 9L, 6L, 6L, 9L, 1L, 6L), .Label = c("ASCUS", "LSIL",
"ASCUS or SIL", "ASC-R", "ASC-R/US", "ASC-US", "ASC-H", "AGC/AGUS",
"BORDERLINE DYSKARYOSIS"), class = "factor"), num = c(384, 48,
23, 61, 105, 31, 1306, 74, 138, 306, 32, 12, 128, 67, 25, 121,
23, 109, 21, 2501, 49, 835, 37, 35, 930, 34, 1680, 266, 438,
115, 238, 542), denom = c(973, 111, 57, 195, 360, 75, 2301, 213,
270, 949, 123, 23, 278, 148, 93, 275, 52, 156, 92, 7334, 66,
1880, 51, 90, 2319, 71, 3681, 672, 1285, 190, 757, 1171), frac = c(0.394699990749359,
0.432399988174438, 0.403499990701675, 0.312799990177155, 0.291700005531311,
0.413300007581711, 0.567600011825562, 0.347400009632111, 0.511099994182587,
0.322400003671646, 0.260199993848801, 0.521700024604797, 0.460399985313416,
0.452699989080429, 0.268799990415573, 0.439999997615814, 0.442299991846085,
0.69870001077652, 0.228300005197525, 0.340999990701675, 0.742399990558624,
0.444099992513657, 0.725499987602234, 0.388900011777878, 0.400999993085861,
0.478899985551834, 0.456400007009506, 0.395799994468689, 0.340856045484543,
0.605300009250641, 0.314399987459183, 0.462900012731552), se = c(0.0156999994069338,
0.0469999983906746, 0.0649999976158142, 0.0331999994814396, 0.0240000002086163,
0.0568999983370304, 0.0103000001981854, 0.032600000500679, 0.0304000005125999,
0.0152000002563, 0.0395999997854233, 0.10419999808073, 0.029899999499321,
0.0408999994397163, 0.046000000089407, 0.029899999499321, 0.0688999965786934,
0.0366999991238117, 0.0438000001013279, 0.00549999997019768,
0.0538000017404556, 0.0115000000223517, 0.0625, 0.0513999983668327,
0.0102000003680587, 0.0593000017106533, 0.00820000004023314,
0.0188999995589256, 0.0132228191941977, 0.0355000011622906, 0.0168999992311001,
0.0146000003442168), up = c(0.42616218328476, 0.529843688011169,
0.54178661108017, 0.382952570915222, 0.341589659452438, 0.532972872257233,
0.587941765785217, 0.415492951869965, 0.572178602218628, 0.353226482868195,
0.346979528665543, 0.731803834438324, 0.520975351333618, 0.536525011062622,
0.370758354663849, 0.500865280628204, 0.586724460124969, 0.769499897956848,
0.327510416507721, 0.351993471384048, 0.842233598232269, 0.46694752573967,
0.841072738170624, 0.497442901134491, 0.421312838792801, 0.600784838199615,
0.47265362739563, 0.433942914009094, 0.367502719163895, 0.675257563591003,
0.348808646202087, 0.491900950670242), lo = c(0.363788783550262,
0.338717311620712, 0.275612711906433, 0.248484954237938, 0.245208755135536,
0.300753623247147, 0.547044515609741, 0.283657312393188, 0.449797093868256,
0.292769432067871, 0.185249075293541, 0.305878013372421, 0.40074035525322,
0.370811879634857, 0.182117596268654, 0.380441725254059, 0.304695636034012,
0.62021142244339, 0.147192806005478, 0.330161929130554, 0.619938731193542,
0.421524852514267, 0.582552492618561, 0.287862300872803, 0.381007432937622,
0.358779191970825, 0.440211117267609, 0.358642756938934, 0.314939588308334,
0.531930029392242, 0.281442880630493, 0.433990597724915)), datalabel = "", time.stamp = "15 Jan 2014 14:08", .Names = c("study",
"author", "year", "tgroup", "num", "denom", "frac", "se", "up",
"lo"), formats = c("%20s", "%14s", "%8.0g", "%22.0g", "%9.0g",
"%9.0g", "%9.0g", "%9.0g", "%9.0g", "%9.0g"), types = c(20L,
14L, 252L, 251L, 254L, 254L, 254L, 254L, 254L, 254L), val.labels = c("",
"", "", "tgroup", "", "", "", "", "", ""), var.labels = c("",
"", "", "", "", "", "", "", "Upper limit", "Lower limit"), row.names = c("1",
"2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13",
"14", "15", "16", "17", "18", "19", "20", "21", "22", "23", "24",
"25", "26", "27", "28", "29", "30", "31", "32"), version = 12L, label.table = structure(list(
tgroup = structure(c(1L, 2L, 3L, 11L, 12L, 13L, 14L, 15L,
16L), .Names = c("ASCUS", "LSIL", "ASCUS or SIL", "ASC-R",
"ASC-R/US", "ASC-US", "ASC-H", "AGC/AGUS", "BORDERLINE DYSKARYOSIS"
))), .Names = "tgroup"), class = "data.frame")
这里是荟萃分析的代码：
library(meta)
(mtprop <- metaprop(event=num, n=denom, studlab=study, byvar=droplevels(tgroup), data=df))
命令metaprop产生以下结果：
proportion 95%-CI %W(fixed) %W(random) byvar
Manos, 1999 0.3947 [0.3638; 0.4262] 3.8 3.7 1
Bergeron, 2000 0.4324 [0.3387; 0.5298] 0.4 2.9 1
Lytwyn, 2000 0.4035 [0.2756; 0.5418] 0.2 2.4 1
Shlay, 2000 0.3128 [0.2485; 0.3830] 0.7 3.2 1
Morin, 2001 0.2917 [0.2452; 0.3416] 1.2 3.4 1
Rebello, 2001 0.4133 [0.3008; 0.5330] 0.3 2.6 3
Solomon, 2001 0.5676 [0.5470; 0.5879] 9.1 3.7 1
Zielinski, 2001 0.3474 [0.2837; 0.4155] 0.8 3.3 3
Kulasingam, 2002 0.5111 [0.4498; 0.5722] 1.1 3.4 1
Pretorius, 2002 0.3224 [0.2928; 0.3532] 3.4 3.7 1
Cuzick, 2003 0.2602 [0.1852; 0.3470] 0.4 2.8 3
Guyot, 2003 0.5217 [0.3059; 0.7318] 0.1 1.6 3
Lonky, 2003 0.4604 [0.4007; 0.5210] 1.1 3.4 1
Wensveen, 2003 0.4527 [0.3708; 0.5365] 0.6 3.1 1
Bruner, 2004 0.2688 [0.1821; 0.3708] 0.3 2.6 2
Rowe, 2004 0.4400 [0.3804; 0.5009] 1.1 3.4 1
Andersson, 2005 0.4423 [0.3047; 0.5867] 0.2 2.3 1
Dalla Palma, 2005 0.6987 [0.6202; 0.7695] 0.5 3.1 1
Giovannelli, 2005 0.2283 [0.1472; 0.3275] 0.3 2.5 1
Kendall, 2005 0.3410 [0.3302; 0.3520] 26.7 3.8 1
Nieh, 2005 0.7424 [0.6199; 0.8422] 0.2 2.3 1
Bergeron, 2006 0.4441 [0.4215; 0.4669] 7.5 3.7 1
Kelly, 2006 0.7255 [0.5826; 0.8411] 0.2 2.1 2
Kiatpongsan, 2006 0.3889 [0.2879; 0.4974] 0.3 2.8 1
Ko, 2006 0.4010 [0.3810; 0.4213] 9.0 3.7 2
Monsonego, 2006 0.4789 [0.3588; 0.6008] 0.3 2.6 1
Moss, 2006 0.4564 [0.4402; 0.4727] 14.8 3.8 3
Selvaggi, 2006 0.3958 [0.3586; 0.4339] 2.6 3.6 2
Wright, 2006 0.3409 [0.3149; 0.3675] 4.7 3.7 2
Cuschieri, 2007 0.6053 [0.5319; 0.6753] 0.7 3.2 3
Ronco, 2007 0.3144 [0.2814; 0.3488] 2.6 3.6 1
You, 2007 0.4629 [0.4340; 0.4919] 4.7 3.7 2
Number of studies combined: k = 32
proportion 95%-CI z p-value
Fixed effect model 0.4084 [0.4024; 0.4144] -- --
Random effects model 0.4249 [0.3921; 0.4583] -- --
Quantifying heterogeneity:
tau^2 = 0.1262; H = 4.79 [4.33; 5.31]; I^2 = 95.7% [94.7%; 96.5%]
Test of heterogeneity:
Q d.f. p-value
712.74 31 < 0.0001
Results for subgroups (fixed effect model):
k proportion 95%-CI Q tau^2 I^2
byvar = ASCUS 20 0.3990 [0.3914; 0.4067] 562.93 0.1879 96.6%
byvar = ASC-US 6 0.4007 [0.3879; 0.4137] 63.59 0.062 92.1%
byvar = BORDERLINE DYSKARYOSIS 6 0.4524 [0.4375; 0.4674] 45.10 0.1551 88.9%
Test for subgroup differences (fixed effect model):
Q d.f. p-value
Between groups 41.13 2 < 0.0001
Within groups 671.62 29 < 0.0001
Results for subgroups (random effects model):
k proportion 95%-CI Q tau^2 I^2
byvar = ASCUS 20 0.4281 [0.3793; 0.4783] 562.93 0.1879 96.6%
byvar = ASC-US 6 0.4106 [0.3574; 0.4660] 63.59 0.062 92.1%
byvar = BORDERLINE DYSKARYOSIS 6 0.4271 [0.3434; 0.5151] 45.10 0.1551 88.9%
Test for subgroup differences (random effects model):
Q d.f. p-value
Between groups 0.24 2 0.8891
Details on meta-analytical method:
- Inverse variance method
- DerSimonian-Laird estimator for tau^2
- Logit transformation
- Clopper-Pearson confidence interval for individual studies
森林情节是：
forest(mtprop)
答案 1 :(得分：0)
这里是我的代码 我使用了元数据包中包含的Olkin95数据的控制臂（仅适用于20个研究）。 定义了小组，将年份分成三组，每组相等。 我决定只将随机效应模型的结果放在图中（comb.fixed = FALSE）
data(Olkin95)
meta <- metaprop(event = event.c,
n = n.c, studlab = paste(author, year),
byvar = cut(year, 3),
data = Olkin95[1:20,])
forest(meta, comb.fixed = FALSE,
bylab = "Years subgroup",
hetlab = "", print.tau2 = FALSE,
layout = "RevMan",
col.square = "black",
col.square.lines = "black")
在meta包中，森林图可以以不同的布局呈现，包括Cochrane协作所使用的布局（layout =“ RevMan”）。 森林功能使用网格图形系统，这很棒，但也有其缺点。我发现不可能通过par（mfrow（））在一个面板中放置不同的森林图，因此我需要使用grid.grab函数捕获一个森林图，并使用grid.arrange将森林图放置在一个面板中。 代码的骨架如下所示：
f1 <- metaprop()
forest(f1)
p1 <- grid.grab()
f2 <- metaprop()
forest(f2)
p2 <- grid.grab()
grid.newpage()
grid.arrange(p1, p2, ncol = 1)
相关问题
最新问题
- 我写了这段代码，但我无法理解我的错误
- 我无法从一个代码实例的列表中删除 None 值，但我可以在另一个实例中。为什么它适用于一个细分市场而不适用于另一个细分市场？
- 是否有可能使 loadstring 不可能等于打印？卢阿
- java中的random.expovariate()
- Appscript 通过会议在 Google 日历中发送电子邮件和创建活动
- 为什么我的 Onclick 箭头功能在 React 中不起作用？
- 在此代码中是否有使用“this”的替代方法？
- 在 SQL Server 和 PostgreSQL 上查询，我如何从第一个表获得第二个表的可视化
- 每千个数字得到
- 更新了城市边界 KML 文件的来源？