2.1 Literature search results
A total of 418 articles were obtained through preliminary search of the database, 320 articles were included after deduplication, 268 articles were initially included after reading the title and abstract, and then the full text was read and 96 articles were excluded according to the established literature inclusion and exclusion criteria, and 18 articles were finally included There are a total of 2668 patients in the literature (Figure 1). Among them, there are 13 articles evaluating the diagnostic value of CRISPR technology [10-22], and 4 articles on chest CT examination [23-26].
2.2 Basic information of the included studies
Table 1 and Table 2are the basic information of all the included documents.
2.3 Meta analysis results
2.3.1 Threshold effect analysis
By drawing the SROC curve, there is no “shoulder-arm-like” distribution, which shows that there is no threshold effect in this study.
2.3.2 Non-threshold effect heterogeneity analysis
For all indicators I2>50%, a random effects model was used for Meta analysis. The results showed that the combined sensitivity of CRISPR technology for diagnosing new coronary pneumonia infection was 0.96 [95% CI (0.93, 0.98)], and the combined specificity was 1.00 [95% CI ( 0.92, 1.00)], the combined positive likelihood ratio is 458.69 [95% CI (11.51, 18280.8)], the combined negative likelihood ratio is 0.04 [95% CI (0.02, 0.07)], the area under the SROC curve is 0.99 [95 %CI (0.97, 0.99)]. The combined sensitivity of chest CT in diagnosing new coronary pneumonia infection was 0.94 [95% CI (0.83, 0.98)], combined specificity was 0.55 [95% CI (0.22, 0.83)], combined diagnostic odds ratio was 19.90 [95% CI (7.88) , 50.25)], the combined positive likelihood ratio is 2.08 [95% CI (1.00, 4.32)], the combined negative likelihood ratio is 0.10 [95% CI (0.05, 0.23)], and the area under the SROC curve is 0.91 [95% CI (0.88, 0.93)] See Figure 3~Figure 9.
2.3.3 Subgroup analysis
In view of the certain heterogeneity of this study, subgroup analysis was performed according to the publication year, country, inspection type, type of CRISPR involved, number of detected target genes, and number of included studies included in the literature study. In terms of CRISPR type, SHERLOCK and non- The pooled sensitivities of SHERLOCK were 0.89 (0.81, 0.95) and 0.97 (0.94, 0.98); pooled specificities were 1.00 (0.95, 1.00) and 1.00 (0.91, 1.00), respectively; pooled DOR was 651.44 (63.19,6716.33), respectively And 10076 (339,299372). The analysis found that the two subgroups did not overlap in specificity, sensitivity, and DOR 95% CI interval in the number of target genes detected, indicating that the difference between the two subgroups was statistically significant; SHERLOCK and non-SHERLOCK detection of AUC They were 0.5 and 0.98, respectively, and the difference in diagnostic accuracy was statistically significant. The results are shown inTable 3.
2.3.4 Meta regression analysis
In view of the certain heterogeneity of this study, subgroup analysis was performed according to the publication year, country, inspection type, type of CRISPR involved, number of detected target genes, and number of included studies included in the literature study. In terms of CRISPR type, SHERLOCK and non- The pooled sensitivities of SHERLOCK were 0.89 (0.81, 0.95) and 0.97 (0.94, 0.98); pooled specificities were 1.00 (0.95, 1.00) and 1.00 (0.91, 1.00), respectively; pooled DOR was 651.44 (63.19,6716.33), respectively And 10076 (339,299372). The analysis found that the two subgroups did not overlap in specificity, sensitivity, and DOR 95% CI interval in the number of target genes detected, indicating that the difference between the two subgroups was statistically significant; SHERLOCK and non-SHERLOCK detection of AUC They were 0.5 and 0.98, respectively, and the difference in diagnostic accuracy was statistically significant. The results are shown in Table 3.
2.4 Sensitivity analysis
This meta-analysis uses the sensitivity analysis method that comes with the RevMan5.4 software to analyze and exclude each group of indicators included in the study. First, single articles are eliminated one by one, and then the analysis is recalculated. Observingthe I2, P value, and the combined effect size and 95% CI, after the single article was eliminated, and finally found that the results did not change much, we can see that the results of this study are reliable.
2.5 Publication bias
The Deek funnel chart was used to test the publication bias, suggesting that there is no potential publication bias among the included studies (P values are 0.03 and 0.55, respectively), as shown in Figure 10and Figure 11.
Table 1 Basic information of the included literature related to the diagnosis of new coronary pneumonia by CRISPR
1st author
|
year
|
country
|
Published journals
|
TP
|
FP
|
FN
|
TN
|
type of CRISPR
|
Target gene
|
XiongErhu[10]
|
2021
|
China
|
Angew. Chem
|
34
|
0
|
1
|
29
|
CRISPR-cas9
|
ORF1ab, E
|
Zhu Xiong[11]
|
2021
|
China
|
ACS sensors
|
37
|
0
|
0
|
77
|
CRISPR-Cas12a
|
ORF1ab, N
|
ShreeyaAgrawal[12]
|
2020
|
United States
|
medRxiv
|
33
|
0
|
2
|
28
|
DISCoVER
|
N
|
ParinazFozouni[13]
|
2021
|
United States
|
Cell
|
5
|
0
|
0
|
5
|
CRISPR-cas13
|
N
|
Dan Xiong[14]
|
2020
|
China
|
medRxiv
|
11
|
0
|
0
|
11
|
CRISPR-Cas12a
|
ORF1ab, N
|
PattarapornNimsamer[15]
|
2020
|
Thailand
|
J Virol Methods
|
41
|
0
|
3
|
63
|
CRISPR-Cas12a
|
N1
|
Jon Arizti-Sanz[16]
|
2020
|
United States
|
bioRxiv
|
27
|
3
|
0
|
20
|
SHINE
|
ORF1a
|
Broughton JP[17]
|
2020
|
United States
|
Nature
|
38
|
0
|
2
|
42
|
DETECTRE
|
N, E
|
MaturadaPatchsung[18]
|
2020
|
Thailand
|
Nat Biomed Eng
|
71
|
0
|
10
|
73
|
SHERLOCK
|
S
|
Jennifer N. Rauch[19]
|
2020
|
United States
|
J Clin Microbiol
|
63
|
3
|
2
|
150
|
CREST
|
N
|
ZahirAli[20]
|
2020
|
Saudi Arabia
|
Virus Research
|
18
|
3
|
0
|
3
|
iSCAN
|
N, E
|
Ashwin Ramachandran[21]
|
2020
|
United States
|
Proc Natl Acad Sci U S A
|
30
|
0
|
2
|
32
|
CRISPR-Cas12a
|
N, E
|
Julia Joung[22]
|
2020
|
United States
|
medRxiv
|
12
|
0
|
0
|
5
|
SHERLOCK
|
N
|
Table 2 Basic information of the included literature related to the diagnosis of new coronary pneumonia by chest CT
1st author |
year |
country |
Published journals |
TP |
FP |
FN |
TN
|
|
Damiano Caruso[23]
|
2020
|
Italy
|
Radiology
|
60
|
42
|
2
|
54
|
chest CT
|
Wu J[24]
|
2020
|
China
|
Invest Radiol
|
76
|
14
|
4
|
6
|
chest CT
|
Yuki Himoto[25]
|
2020
|
Japan
|
Jpn J Radiol
|
4
|
1
|
2
|
14
|
chest CT
|
ZenghuiCheng[26]
|
2020
|
China
|
AJR Am J Roentgenol
|
11
|
17
|
0
|
5
|
chest CT
|
Table 3 CRISPR technology and chest CT diagnosis of new coronary pneumonia infection subgroup analysis results (point estimation and 95% CI)
type of detection
|
subgroup
|
Quantity included
|
sensitivity
|
specificity
|
Diagnostic odds ratio
|
Positive likelihood ratio
|
Negative likelihood ratio
|
SROC
|
|
2020
|
10
|
0.96
|
0.99
|
4452.91
|
183.64
|
0.04
|
0.99
|
(0.91,0.98)
|
(0.90,1.00)
|
(259.99,76265.41)
|
(8.57,3934.04)
|
(0.02,0.09)
|
(0.97,0.99)
|
|
2021
|
3
|
0.99
|
1
|
1308.57
|
44.51
|
0.04
|
0.99
|
(0.93,1.00)
|
(0.97,1.00)
|
(119.59,14318.96)
|
(9, 220.01)
|
(0.01,0.13)
|
(0.96,0.99)
|
|
China
|
3
|
0.96
(0.94,0.97)
|
1
|
1942.19
|
58.19
|
0.03
|
0.99
|
(0.97,1.00)
|
(232.77,16205.42)
|
(11.94,283.51)
|
(0.01,0.11)
|
(0.98,1.00)
|
|
Other countries
|
10
|
0.96
(0.91,0.98)
|
0.99
|
4150.83
|
175.11
|
0.04
|
0.99
|
(0.89,1.00)
|
(246.07,70017.52)
|
(8.27, 3704.79)
|
(0.02,0.09)
|
(0.97,0.99)
|
CRISPR
|
Single gene
|
7
|
0.96
(0.89,0.98)
|
0.26
(0.13,0.42)
|
2552
|
114.4
|
0.04
|
0.99
|
(320, 20376)
|
(11.8, 1109.6)
|
(0.02,0.11)
|
(0.98,1.01)
|
|
Multiple genes
|
6
|
0.97
(0.93,0.99)
|
1
|
767708
|
21110
|
0.03
|
0.98
|
(0.03,1.00)
|
(1, 4.3e+11)
|
(10.0,1.4e+10)
|
(0.01,0.07)
|
(0.96,0.99)
|
|
SHERLOCK
|
2
|
0.89
|
1
|
651.44
|
37.58
|
0.12
|
0.5
|
(0.81,0.95)
|
(0.95,1.00)
|
(63.19,6716.33)
|
(2.83,498.59)
|
(0.07,0.21)
|
|
Non-SHERLOCK
|
11
|
0.97
|
1
|
10076
|
316.1
|
0.03
|
0.98
|
(0.94,0.98)
|
(0.91,1.00)
|
-339,299,372
|
(9.7,10265.5)
|
(0.02,0.06)
|
(0.97,0.99)
|
|
China
|
2
|
0.96
|
0.26
(0.15,0.41)
|
7.79
|
1.3
|
0.17 (0.06,0.49)
|
0.8
|
(0.89,0.99)
|
(2.26,28.07)
|
(1.07,1.58)
|
(0.76,0.83)
|
Chest CT
|
Other countries
|
2
|
0.89
(0.51,0.99)
|
0.78
(0.38,0.95)
|
29.11
|
3.99
|
0.14
|
0.9
|
(6.61,128.12)
|
(1.21,13.12)
|
(0.03,0.69)
|
(0.89,0.93)
|
|
>50
|
2
|
0.96
(0.89,0.99)
|
0.26
|
7.97
|
1.3
|
0.17
|
0.94
|
(0.14,0.42)
|
(2.26,28.07)
|
(1.07,1.58)
|
(0.06,0.49)
|
(0.92,0.98)
|
|
<50
|
2
|
0.88
|
0.51
|
15.45
|
3.11
|
0.32
|
0.5
|
(0.64,0.99)
|
(0.34,0.68)
|
(2.13,111.99)
|
(0.21,46.75)
|
(0.11,0.93)
|