How to conduct AMR analysis
Matthijs S. Berends
-07 June 2019
+15 June 2019
AMR.RmdNote: values on this page will change with every website update since they are based on randomly created values and the page was written in R Markdown. However, the methodology remains unchanged. This page was generated on 07 June 2019.
+Note: values on this page will change with every website update since they are based on randomly created values and the page was written in R Markdown. However, the methodology remains unchanged. This page was generated on 15 June 2019.
Introduction
@@ -224,21 +217,21 @@Dates
-Let’s pretend that our data consists of blood cultures isolates from 1 January 2010 until 1 January 2018.
+Let’s pretend that our data consists of blood cultures isolates from between 1 January 2010 and 1 January 2018.
This dates object now contains all days in our date range.
Using the left_join() function from the dplyr package, we can ‘map’ the gender to the patient ID using the patients_table object we created earlier:
The resulting data set contains 5,000 blood culture isolates. With the head() function we can preview the first 6 values of this data set:
The resulting data set contains 20,000 blood culture isolates. With the head() function we can preview the first 6 values of this data set:
| 2010-01-26 | -S3 | -Hospital B | -Escherichia coli | -R | -I | -R | -S | -F | -||||
| 2010-08-26 | -G7 | -Hospital D | -Escherichia coli | -R | -S | -S | -S | -M | -||||
| 2016-05-04 | -B8 | +2010-06-23 | +R10 | Hospital A | -Escherichia coli | -S | -S | -R | -S | -M | -||
| 2016-11-19 | -E10 | -Hospital D | Streptococcus pneumoniae | S | S | S | S | -M | -||||
| 2013-07-13 | -G3 | -Hospital A | -Escherichia coli | -R | -S | -S | -S | -M | +F | |||
| 2017-05-23 | -V3 | -Hospital C | -Staphylococcus aureus | +2016-04-09 | +R1 | +Hospital D | +Klebsiella pneumoniae | S | S | S | S | F |
| 2017-01-27 | +U10 | +Hospital A | +Streptococcus pneumoniae | +R | +S | +R | +S | +F | +||||
| 2017-05-23 | +C4 | +Hospital B | +Klebsiella pneumoniae | +R | +S | +R | +S | +M | +||||
| 2015-03-27 | +W3 | +Hospital B | +Escherichia coli | +S | +S | +S | +S | +F | +||||
| 2014-06-14 | +L5 | +Hospital A | +Escherichia coli | +S | +S | +S | +S | +M | +
Now, let’s start the cleaning and the analysis!
@@ -418,9 +411,9 @@ # # Item Count Percent Cum. Count Cum. Percent # --- ----- ------- -------- ----------- ------------- -# 1 M 10,368 51.8% 10,368 51.8% -# 2 F 9,632 48.2% 20,000 100.0% -So, we can draw at least two conclusions immediately. From a data scientist perspective, the data looks clean: only values M and F. From a researcher perspective: there are slightly more men. Nothing we didn’t already know.
So, we can draw at least two conclusions immediately. From a data scientists perspective, the data looks clean: only values M and F. From a researchers perspective: there are slightly more men. Nothing we didn’t already know.
The data is already quite clean, but we still need to transform some variables. The bacteria column now consists of text, and we want to add more variables based on microbial IDs later on. So, we will transform this column to valid IDs. The mutate() function of the dplyr package makes this really easy:
So only 28.5% is suitable for resistance analysis! We can now filter on it with the filter() function, also from the dplyr package:
So only 28.3% is suitable for resistance analysis! We can now filter on it with the filter() function, also from the dplyr package:
For future use, the above two syntaxes can be shortened with the filter_first_isolate() function:
First weighted isolates
-We made a slight twist to the CLSI algorithm, to take into account the antimicrobial susceptibility profile. Imagine this data, sorted on date:
+We made a slight twist to the CLSI algorithm, to take into account the antimicrobial susceptibility profile. Have a look at all isolates of patient V3, sorted on date:
| isolate | @@ -536,21 +529,21 @@||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | -2010-01-16 | -V9 | +2010-01-06 | +V3 | B_ESCHR_COL | S | S | S | -S | +R | TRUE | |
| 2 | -2010-02-23 | -V9 | +2010-07-24 | +V3 | B_ESCHR_COL | -S | +R | S | S | S | @@ -558,10 +551,10 @@||
| 3 | -2010-04-05 | -V9 | +2010-07-26 | +V3 | B_ESCHR_COL | -S | +R | S | S | S | @@ -569,19 +562,19 @@||
| 4 | -2010-04-28 | -V9 | +2011-05-06 | +V3 | B_ESCHR_COL | +R | S | S | S | -S | -FALSE | +TRUE |
| 5 | -2010-05-20 | -V9 | +2011-06-04 | +V3 | B_ESCHR_COL | S | S | @@ -591,41 +584,41 @@|||||
| 6 | -2010-07-07 | -V9 | +2011-07-22 | +V3 | B_ESCHR_COL | S | S | +S | +S | +FALSE | +||
| 7 | +2011-08-15 | +V3 | +B_ESCHR_COL | +I | +I | +S | +R | +FALSE | +||||
| 8 | +2011-09-20 | +V3 | +B_ESCHR_COL | +R | +S | R | S | FALSE | ||||
| 7 | -2010-09-11 | -V9 | -B_ESCHR_COL | -S | -S | -S | -S | -FALSE | -||||
| 8 | -2010-09-19 | -V9 | -B_ESCHR_COL | -S | -S | -S | -S | -FALSE | -||||
| 9 | -2010-10-07 | -V9 | +2012-03-26 | +V3 | B_ESCHR_COL | S | S | @@ -635,18 +628,18 @@|||||
| 10 | -2011-01-14 | -V9 | +2012-06-01 | +V3 | B_ESCHR_COL | R | +I | S | S | -S | -FALSE | +TRUE |
Only 1 isolates are marked as ‘first’ according to CLSI guideline. But when reviewing the antibiogram, it is obvious that some isolates are absolutely different strains and should be included too. This is why we weigh isolates, based on their antibiogram. The key_antibiotics() function adds a vector with 18 key antibiotics: 6 broad spectrum ones, 6 small spectrum for Gram negatives and 6 small spectrum for Gram positives. These can be defined by the user.
Only 3 isolates are marked as ‘first’ according to CLSI guideline. But when reviewing the antibiogram, it is obvious that some isolates are absolutely different strains and should be included too. This is why we weigh isolates, based on their antibiogram. The key_antibiotics() function adds a vector with 18 key antibiotics: 6 broad spectrum ones, 6 small spectrum for Gram negatives and 6 small spectrum for Gram positives. These can be defined by the user.
If a column exists with a name like ‘key(…)ab’ the first_isolate() function will automatically use it and determine the first weighted isolates. Mind the NOTEs in below output:
data <- data %>%
mutate(keyab = key_antibiotics(.)) %>%
@@ -657,7 +650,7 @@
# NOTE: Using column `patient_id` as input for `col_patient_id`.
# NOTE: Using column `keyab` as input for `col_keyantibiotics`. Use col_keyantibiotics = FALSE to prevent this.
# [Criterion] Inclusion based on key antibiotics, ignoring I.
-# => Found 15,072 first weighted isolates (75.4% of total)| isolate | @@ -674,34 +667,34 @@||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | -2010-01-16 | -V9 | +2010-01-06 | +V3 | B_ESCHR_COL | S | S | S | -S | +R | TRUE | TRUE | ||
| 2 | -2010-02-23 | -V9 | +2010-07-24 | +V3 | B_ESCHR_COL | -S | +R | S | S | S | FALSE | -FALSE | +TRUE | |
| 3 | -2010-04-05 | -V9 | +2010-07-26 | +V3 | B_ESCHR_COL | -S | +R | S | S | S | @@ -710,95 +703,95 @@||||
| 4 | -2010-04-28 | -V9 | +2011-05-06 | +V3 | B_ESCHR_COL | +R | S | S | S | -S | -FALSE | -FALSE | +TRUE | +TRUE |
| 5 | -2010-05-20 | -V9 | +2011-06-04 | +V3 | B_ESCHR_COL | S | S | S | S | FALSE | -FALSE | -|||
| 6 | -2010-07-07 | -V9 | -B_ESCHR_COL | -S | -S | -R | -S | -FALSE | TRUE | |||||
| 7 | -2010-09-11 | -V9 | +||||||||||||
| 6 | +2011-07-22 | +V3 | B_ESCHR_COL | S | S | S | S | FALSE | +FALSE | +|||||
| 7 | +2011-08-15 | +V3 | +B_ESCHR_COL | +I | +I | +S | +R | +FALSE | TRUE | |||||
| 8 | -2010-09-19 | -V9 | +2011-09-20 | +V3 | B_ESCHR_COL | +R | S | -S | -S | +R | S | FALSE | -FALSE | +TRUE |
| 9 | -2010-10-07 | -V9 | +2012-03-26 | +V3 | B_ESCHR_COL | S | S | S | S | FALSE | -FALSE | -|||
| 10 | -2011-01-14 | -V9 | -B_ESCHR_COL | -R | -S | -S | -S | -FALSE | +TRUE | +|||||
| 10 | +2012-06-01 | +V3 | +B_ESCHR_COL | +R | +I | +S | +S | +TRUE | TRUE |
Instead of 1, now 4 isolates are flagged. In total, 75.4% of all isolates are marked ‘first weighted’ - 46.9% more than when using the CLSI guideline. In real life, this novel algorithm will yield 5-10% more isolates than the classic CLSI guideline.
+Instead of 3, now 8 isolates are flagged. In total, 76% of all isolates are marked ‘first weighted’ - 47.7% more than when using the CLSI guideline. In real life, this novel algorithm will yield 5-10% more isolates than the classic CLSI guideline.
As with filter_first_isolate(), there’s a shortcut for this new algorithm too:
So we end up with 15,072 isolates for analysis.
+So we end up with 15,202 isolates for analysis.
We can remove unneeded columns:
@@ -806,7 +799,6 @@| date | patient_id | hospital | @@ -823,89 +815,68 @@|||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | -2010-01-26 | -S3 | -Hospital B | -B_ESCHR_COL | -R | -I | -R | +2010-06-23 | +R10 | +Hospital A | +B_STRPT_PNE | S | +S | +S | +R | F | -Gram negative | -Escherichia | -coli | +Gram-positive | +Streptococcus | +pneumoniae | TRUE |
| 2 | -2010-08-26 | -G7 | +2016-04-09 | +R1 | Hospital D | -B_ESCHR_COL | +B_KLBSL_PNE | R | S | S | S | -M | -Gram negative | -Escherichia | -coli | -TRUE | -|||||||
| 3 | -2016-05-04 | -B8 | -Hospital A | -B_ESCHR_COL | -S | -S | -R | -S | -M | -Gram negative | -Escherichia | -coli | -TRUE | -||||||||||
| 5 | -2013-07-13 | -G3 | -Hospital A | -B_ESCHR_COL | -R | -S | -S | -S | -M | -Gram negative | -Escherichia | -coli | -TRUE | -||||||||||
| 8 | -2016-11-20 | -Y6 | -Hospital A | -B_ESCHR_COL | -S | -S | -R | -S | F | -Gram negative | -Escherichia | -coli | +Gram-negative | +Klebsiella | +pneumoniae | +TRUE | +|||||||
| 2017-01-27 | +U10 | +Hospital A | +B_STRPT_PNE | +R | +R | +R | +R | +F | +Gram-positive | +Streptococcus | +pneumoniae | TRUE | |||||||||||
| 10 | -2015-11-03 | -W8 | +2017-05-23 | +C4 | +Hospital B | +B_KLBSL_PNE | +R | +S | +R | +S | +M | +Gram-negative | +Klebsiella | +pneumoniae | +TRUE | +||||||||
| 2015-03-27 | +W3 | Hospital B | B_ESCHR_COL | S | @@ -913,7 +884,22 @@S | S | F | -Gram negative | +Gram-negative | +Escherichia | +coli | +TRUE | +|||||||||||
| 2014-06-14 | +L5 | +Hospital A | +B_ESCHR_COL | +S | +S | +S | +S | +M | +Gram-negative | Escherichia | coli | TRUE | @@ -935,9 +921,9 @@|||||||||||
| 1 | Escherichia coli | -7,451 | -49.4% | -7,451 | -49.4% | +7,485 | +49.2% | +7,485 | +49.2% | ||||||||||||||
| 2 | Staphylococcus aureus | -3,713 | -24.6% | -11,164 | -74.1% | +3,758 | +24.7% | +11,243 | +74.0% | ||||||||||||||
| 3 | Streptococcus pneumoniae | -2,367 | -15.7% | -13,531 | -89.8% | +2,371 | +15.6% | +13,614 | +89.6% | ||||||||||||||
| 4 | Klebsiella pneumoniae | -1,541 | -10.2% | -15,072 | +1,588 | +10.4% | +15,202 | 100.0% | |||||||||||||||
| Hospital A | -0.4472198 | +0.4688237 | |||||||||||||||||||||
| Hospital B | -0.4684564 | +0.4693374 | |||||||||||||||||||||
| Hospital C | -0.4662494 | +0.4641460 | |||||||||||||||||||||
| Hospital D | -0.4704907 | +0.4664644 |
Omit the translate_ab = FALSE to have the antibiotic codes (AMX, AMC, CIP, GEN) translated to official WHO names (amoxicillin, amoxicillin and betalactamase inhibitor, ciprofloxacin, gentamicin).
Omit the translate_ab = FALSE to have the antibiotic codes (AMX, AMC, CIP, GEN) translated to official WHO names (amoxicillin, amoxicillin/clavulanic acid, ciprofloxacin, gentamicin).
If we group on e.g. the genus column and add some additional functions from our package, we can create this:
# group the data on `genus`
ggplot(data_1st %>% group_by(genus)) +
@@ -1136,7 +1122,7 @@ Longest: 24
# of which we have 4 (earlier created with `as.rsi`)
geom_rsi(x = "genus") +
# split plots on antibiotic
- facet_rsi(facet = "Antibiotic") +
+ facet_rsi(facet = "antibiotic") +
# make R red, I yellow and S green
scale_rsi_colours() +
# show percentages on y axis
@@ -1154,7 +1140,7 @@ Longest: 24
data_1st %>%
group_by(genus) %>%
ggplot_rsi(x = "genus",
- facet = "Antibiotic",
+ facet = "antibiotic",
breaks = 0:4 * 25,
datalabels = FALSE) +
coord_flip()
@@ -1164,48 +1150,35 @@ Longest: 24
Independence test
The next example uses the included septic_patients, which is an anonymised data set containing 2,000 microbial blood culture isolates with their full antibiograms found in septic patients in 4 different hospitals in the Netherlands, between 2001 and 2017. It is true, genuine data. This data.frame can be used to practice AMR analysis.
-We will compare the resistance to fosfomycin (column FOS) in hospital A and D. The input for the final fisher.test() will be this:
-
-
-A
-D
-
-
-
-IR
-25
-77
-
-
-S
-24
-33
-
-
-
-We can transform the data and apply the test in only a couple of lines:
-septic_patients %>%
+We will compare the resistance to fosfomycin (column FOS) in hospital A and D. The input for the fisher.test() can be retrieved with a transformation like this:
+check_FOS <- septic_patients %>%
filter(hospital_id %in% c("A", "D")) %>% # filter on only hospitals A and D
select(hospital_id, FOS) %>% # select the hospitals and fosfomycin
group_by(hospital_id) %>% # group on the hospitals
- count_df(combine_IR = TRUE) %>% # count all isolates per group (hospital_id)
- tidyr::spread(hospital_id, Value) %>% # transform output so A and D are columns
+ count_df(combine_SI = TRUE) %>% # count all isolates per group (hospital_id)
+ tidyr::spread(hospital_id, value) %>% # transform output so A and D are columns
select(A, D) %>% # and select these only
- as.matrix() %>% # transform to good old matrix for fisher.test()
- fisher.test() # do Fisher's Exact Test
-#
-# Fisher's Exact Test for Count Data
-#
-# data: .
-# p-value = 0.03104
-# alternative hypothesis: true odds ratio is not equal to 1
-# 95 percent confidence interval:
-# 0.2111489 0.9485124
-# sample estimates:
-# odds ratio
-# 0.4488318
-As can be seen, the p value is 0.03, which means that the fosfomycin resistances found in hospital A and D are really different.
+ as.matrix() # transform to good old matrix for fisher.test()
+
+check_FOS
+# A D
+# [1,] 25 77
+# [2,] 24 33
+We can apply the test now with:
+# do Fisher's Exact Test
+fisher.test(check_FOS)
+#
+# Fisher's Exact Test for Count Data
+#
+# data: check_FOS
+# p-value = 0.03104
+# alternative hypothesis: true odds ratio is not equal to 1
+# 95 percent confidence interval:
+# 0.2111489 0.9485124
+# sample estimates:
+# odds ratio
+# 0.4488318
+As can be seen, the p value is 0.031, which means that the fosfomycin resistances found in hospital A and D are really different.
-AMR 0.7.0.9009 Unreleased +AMR 0.7.0.9010 Unreleased
New
-
-
Function
+rsi_df()to transform adata.frameto a data set containing only the microbial interpretation (S, I, R), the antibiotic, the percentage of S/I/R and the number of available isolates. This is a convenient combinations of existing functionscount_df()andportion_df()to immediately show resistance percentages and number of available isolates:Function
rsi_df()to transform adata.frameto a data set containing only the microbial interpretation (S, I, R), the antibiotic, the percentage of S/I/R and the number of available isolates. This is a convenient combination of the existing functionscount_df()andportion_df()to immediately show resistance percentages and number of available isolates: -
-
Support for all scientifically published pathotypes of E. coli to date. Supported are: AIEC (Adherent-Invasive E. coli), ATEC (Atypical Entero-pathogenic E. coli), DAEC (Diffusely Adhering E. coli), EAEC (Entero-Aggresive E. coli), EHEC (Entero-Haemorrhagic E. coli), EIEC (Entero-Invasive E. coli), EPEC (Entero-Pathogenic E. coli), ETEC (Entero-Toxigenic E. coli), NMEC (Neonatal Meningitis‐causing E. coli), STEC (Shiga-toxin producing E. coli) and UPEC (Uropathogenic E. coli). All these lead to the microbial ID of E. coli:
+Support for all scientifically published pathotypes of E. coli to date. Supported are:
+-
+
- AIEC (Adherent-Invasive E. coli) +
- ATEC (Atypical Entero-pathogenic E. coli) +
- DAEC (Diffusely Adhering E. coli) +
- EAEC (Entero-Aggresive E. coli) +
- EHEC (Entero-Haemorrhagic E. coli) +
- EIEC (Entero-Invasive E. coli) +
- EPEC (Entero-Pathogenic E. coli) +
- ETEC (Entero-Toxigenic E. coli) +
- NMEC (Neonatal Meningitis‐causing E. coli) +
- STEC (Shiga-toxin producing E. coli) +
- UPEC (Uropathogenic E. coli) +
All these lead to the microbial ID of E. coli:
+mo_name("UPEC") +# "Escherichia coli" +mo_gramstain("EHEC") +# "Gram-negative"
Function mo_info() as an analogy to ab_info(). The mo_info() prints a list with the full taxonomy, authors, and the URL to the online database of a microorganism
mo_info() as an analogy to ab_info(). The mo_info() prints a list with the full taxonomy, authors, and the URL to the online database of a microorganismFunction mo_synonyms() to get all previously accepted taxonomic names of a microorganism
as.mo()
plot() and barplot() for MIC and RSI classesas.mo(..., allow_uncertain = 3)
Contents