These functions determine which items in a vector can be considered (the start of) a new episode, based on the argument episode_days. This can be used to determine clinical episodes for any epidemiological analysis. The get_episode() function returns the index number of the episode per group, while the is_new_episode() function returns values TRUE/FALSE to indicate whether an item in a vector is the start of a new episode.
Arguments
- x
vector of dates (class
DateorPOSIXt), will be sorted internally to determine episodes- episode_days
required episode length in days, can also be less than a day or
Inf, see Details- ...
ignored, only in place to allow future extensions
Details
Dates are first sorted from old to new. The oldest date will mark the start of the first episode. After this date, the next date will be marked that is at least episode_days days later than the start of the first episode. From that second marked date on, the next date will be marked that is at least episode_days days later than the start of the second episode which will be the start of the third episode, and so on. Before the vector is being returned, the original order will be restored.
The first_isolate() function is a wrapper around the is_new_episode() function, but is more efficient for data sets containing microorganism codes or names and allows for different isolate selection methods.
The dplyr package is not required for these functions to work, but these functions do support variable grouping and work conveniently inside dplyr verbs such as filter(), mutate() and summarise().
Examples
# `example_isolates` is a data set available in the AMR package.
# See ?example_isolates
df <- example_isolates[sample(seq_len(2000), size = 200), ]
get_episode(df$date, episode_days = 60) # indices
#> [1] 47 52 21 26 19 32 19 8 51 55 23 43 55 7 6 20 57 21 45 24 53 34 5 16 15
#> [26] 11 39 26 10 3 13 23 41 29 18 60 4 57 37 63 38 42 15 7 13 60 32 14 50 24
#> [51] 10 1 21 52 1 9 27 49 43 40 4 26 43 56 16 28 19 36 20 13 59 54 51 43 64
#> [76] 52 57 49 46 61 29 13 38 55 16 60 15 26 30 57 21 14 44 28 22 61 56 19 28 25
#> [101] 51 34 37 5 37 31 16 2 6 2 55 46 5 6 23 55 10 61 13 20 1 16 58 64 30
#> [126] 43 26 10 1 30 21 62 12 14 10 55 16 64 46 6 8 53 48 57 43 35 33 27 52 56
#> [151] 57 56 33 32 13 20 26 11 24 6 8 61 5 25 37 42 7 19 58 32 28 9 17 63 62
#> [176] 58 22 18 39 34 34 64 5 42 63 24 52 13 43 3 24 1 34 56 20 62 14 16 32 56
is_new_episode(df$date, episode_days = 60) # TRUE/FALSE
#> [1] TRUE TRUE TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE
#> [13] TRUE TRUE FALSE FALSE FALSE FALSE TRUE FALSE TRUE FALSE FALSE FALSE
#> [25] FALSE TRUE FALSE FALSE FALSE TRUE FALSE FALSE TRUE FALSE FALSE FALSE
#> [37] FALSE TRUE TRUE FALSE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE
#> [49] TRUE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE TRUE FALSE TRUE
#> [61] TRUE TRUE TRUE FALSE FALSE FALSE FALSE TRUE FALSE FALSE TRUE TRUE
#> [73] FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE TRUE TRUE TRUE FALSE
#> [85] FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE TRUE FALSE TRUE TRUE
#> [97] FALSE TRUE TRUE FALSE FALSE TRUE FALSE FALSE FALSE TRUE FALSE TRUE
#> [109] FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE
#> [121] FALSE FALSE FALSE TRUE TRUE FALSE FALSE FALSE FALSE FALSE FALSE TRUE
#> [133] TRUE FALSE FALSE FALSE FALSE FALSE FALSE TRUE FALSE FALSE TRUE FALSE
#> [145] FALSE TRUE FALSE TRUE FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE
#> [157] FALSE FALSE FALSE FALSE TRUE FALSE TRUE TRUE FALSE FALSE FALSE FALSE
#> [169] TRUE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE TRUE TRUE FALSE
#> [181] FALSE FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE FALSE TRUE TRUE
#> [193] FALSE TRUE TRUE FALSE TRUE TRUE FALSE FALSE
# filter on results from the third 60-day episode only, using base R
df[which(get_episode(df$date, 60) == 3), ]
#> # A tibble: 2 × 46
#> date patient age gender ward mo PEN OXA FLC AMX
#> <date> <chr> <dbl> <chr> <chr> <mo> <rsi> <rsi> <rsi> <rsi>
#> 1 2002-11-04 304347 62 M Clinical B_STRPT_PNMN S NA NA S
#> 2 2002-11-11 D80753 74 F Outpatie… B_STPHY_CONS R NA R R
#> # … with 36 more variables: AMC <rsi>, AMP <rsi>, TZP <rsi>, CZO <rsi>,
#> # FEP <rsi>, CXM <rsi>, FOX <rsi>, CTX <rsi>, CAZ <rsi>, CRO <rsi>,
#> # GEN <rsi>, TOB <rsi>, AMK <rsi>, KAN <rsi>, TMP <rsi>, SXT <rsi>,
#> # NIT <rsi>, FOS <rsi>, LNZ <rsi>, CIP <rsi>, MFX <rsi>, VAN <rsi>,
#> # TEC <rsi>, TCY <rsi>, TGC <rsi>, DOX <rsi>, ERY <rsi>, CLI <rsi>,
#> # AZM <rsi>, IPM <rsi>, MEM <rsi>, MTR <rsi>, CHL <rsi>, COL <rsi>,
#> # MUP <rsi>, RIF <rsi>
# the functions also work for less than a day, e.g. to include one per hour:
get_episode(c(
Sys.time(),
Sys.time() + 60 * 60
),
episode_days = 1 / 24
)
#> [1] 1 2
# \donttest{
if (require("dplyr")) {
# is_new_episode() can also be used in dplyr verbs to determine patient
# episodes based on any (combination of) grouping variables:
df %>%
mutate(condition = sample(
x = c("A", "B", "C"),
size = 200,
replace = TRUE
)) %>%
group_by(condition) %>%
mutate(new_episode = is_new_episode(date, 365)) %>%
select(patient, date, condition, new_episode)
}
#> # A tibble: 200 × 4
#> # Groups: condition [3]
#> patient date condition new_episode
#> <chr> <date> <chr> <lgl>
#> 1 B43936 2013-05-16 A FALSE
#> 2 AC8303 2014-10-20 B TRUE
#> 3 E19253 2006-08-22 C FALSE
#> 4 D81577 2008-05-12 C FALSE
#> 5 FD439C 2006-03-18 C FALSE
#> 6 B26404 2010-01-08 C FALSE
#> 7 D99170 2006-03-13 B FALSE
#> 8 955371 2003-12-12 B FALSE
#> 9 C40150 2014-03-05 C TRUE
#> 10 557266 2015-10-12 B FALSE
#> # … with 190 more rows
if (require("dplyr")) {
df %>%
group_by(ward, patient) %>%
transmute(date,
patient,
new_index = get_episode(date, 60),
new_logical = is_new_episode(date, 60)
)
}
#> # A tibble: 200 × 5
#> # Groups: ward, patient [186]
#> ward date patient new_index new_logical
#> <chr> <date> <chr> <dbl> <lgl>
#> 1 ICU 2013-05-16 B43936 1 TRUE
#> 2 Clinical 2014-10-20 AC8303 1 TRUE
#> 3 Clinical 2006-08-22 E19253 1 TRUE
#> 4 Clinical 2008-05-12 D81577 1 TRUE
#> 5 Clinical 2006-03-18 FD439C 1 TRUE
#> 6 Clinical 2010-01-08 B26404 1 TRUE
#> 7 ICU 2006-03-13 D99170 1 TRUE
#> 8 ICU 2003-12-12 955371 1 TRUE
#> 9 Clinical 2014-03-05 C40150 1 TRUE
#> 10 Clinical 2015-10-12 557266 1 TRUE
#> # … with 190 more rows
if (require("dplyr")) {
df %>%
group_by(ward) %>%
summarise(
n_patients = n_distinct(patient),
n_episodes_365 = sum(is_new_episode(date, episode_days = 365)),
n_episodes_60 = sum(is_new_episode(date, episode_days = 60)),
n_episodes_30 = sum(is_new_episode(date, episode_days = 30))
)
}
#> # A tibble: 3 × 5
#> ward n_patients n_episodes_365 n_episodes_60 n_episodes_30
#> <chr> <int> <int> <int> <int>
#> 1 Clinical 110 15 55 73
#> 2 ICU 59 14 33 42
#> 3 Outpatient 17 10 15 15
if (require("dplyr")) {
# grouping on patients and microorganisms leads to the same
# results as first_isolate() when using 'episode-based':
x <- df %>%
filter_first_isolate(
include_unknown = TRUE,
method = "episode-based"
)
y <- df %>%
group_by(patient, mo) %>%
filter(is_new_episode(date, 365)) %>%
ungroup()
identical(x, y)
}
#> Including isolates from ICU.
#> [1] TRUE
if (require("dplyr")) {
# but is_new_episode() has a lot more flexibility than first_isolate(),
# since you can now group on anything that seems relevant:
df %>%
group_by(patient, mo, ward) %>%
mutate(flag_episode = is_new_episode(date, 365)) %>%
select(group_vars(.), flag_episode)
}
#> # A tibble: 200 × 4
#> # Groups: patient, mo, ward [193]
#> patient mo ward flag_episode
#> <chr> <mo> <chr> <lgl>
#> 1 B43936 B_CRYNB ICU TRUE
#> 2 AC8303 B_STPHY_HMLY Clinical TRUE
#> 3 E19253 B_STPHY_CONS Clinical TRUE
#> 4 D81577 B_HMPHL_INFL Clinical TRUE
#> 5 FD439C B_STPHY_CONS Clinical TRUE
#> 6 B26404 B_STPHY_EPDR Clinical TRUE
#> 7 D99170 B_STPHY_EPDR ICU TRUE
#> 8 955371 B_STPHY_AURS ICU TRUE
#> 9 C40150 B_CTBCTR_ACNS Clinical TRUE
#> 10 557266 B_STPHY_HMLY Clinical TRUE
#> # … with 190 more rows
# }