Ko zhene ich, G igo y; Zande , Anne
A icle — Published Ve sion
Le e aging he po en ial o he Ge man ope a ing oom
benchma king ini ia i e o planning: A eady- o-use
su gical p ocess da a se
Heal h Ca e Managemen Science
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Ko zhene ich, G igo y; Zande , Anne (2024) : Le e aging he po en ial o he
Ge man ope a ing oom benchma king ini ia i e o planning: A eady- o-use su gical p ocess da a
se , Heal h Ca e Managemen Science, ISSN 1572-9389, Sp inge US, New Yo k, NY, Vol. 27, Iss. 3,
pp. 328-351,
h ps://doi.o g/10.1007/s10729-024-09672-9
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/315267
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Heal h Ca e Managemen Science (2024) 27:328–351
h ps://doi.o g/10.1007/s10729-024-09672-9
Le e aging he po en ial o he Ge man ope a ing oom benchma king
ini ia i e o planning: A eady- o-use su gical p ocess da a se
G igo y Ko zhene ich1·Anne Zande 2
Recei ed: 31 Decembe 2022 / Accep ed: 13 Ap il 2024 / Published online: 2 May 2024
© The Au ho (s) 2024
Abs ac
We p esen a eely a ailable da a se o su gical case mixes and su ge y p ocess du a ion dis ibu ions based on p ocessed
da a om he Ge man Ope a ing Room Benchma king ini ia i e. This ini ia i e collec s su gical p ocess da a om o e
320 Ge man, Aus ian, and Swiss hospi als. The da a exhibi s high le els o quan i y, quali y, s anda diza ion, and mul i-
dimensionali y, making i especially aluable o ope a ing oom planning in Ope a ions Resea ch. We conside de ailed s eps
o he pe iope a i e p ocess and g oup he da a wi h espec o he hospi al’s le el o ca e, he su ge y special y, and he ype
o su ge y pa ien . We compa e case mixes o di e en subg oups and conclude ha hey di e signi ican ly, demons a ing
ha i is necessa y o es ope a ing oom planning me hods in di e en se ings, e.g., using da a se s like ou s. Fu he , we
discuss limi a ions and u u e esea ch di ec ions. Finally, we encou age he ex ension and ounda ion o new ope a ing oom
benchma king ini ia i es and hei usage o ope a ing oom planning.
Keywo ds Su gical P ocess Da a ·Da a analysis ·Ge man Pe iope a i e P ocedu al Time Glossa y ·Ope a ing Room
Benchma king Ini ia i e ·Ope a ing Room Planning ·Ope a ions Resea ch
Highligh s
•We show he sui abili y o he su ge y p ocess da a (wi h
high le els o quan i y and quali y, s anda diza ion, and
mul i-dimensionali y) om he Ge man Ope a ing Room
Benchma king ini ia i e o ope a ing oom planning.
•We p esen a p ocessed da a se o case mixes and
de ailed su ge y p ocess du a ion dis ibu ions g ouped
wi h espec o hospi al le el o ca e, su gical special y,
and ype o su gical pa ien .
•We make he p ocessed da a se eely a ailable o
esea che s wo king on ope a ing oom planning.
G igo y Ko zhene ich and Anne Zande con ibu ed equally o his wo k.
BAnne Zande
a.b[email p o ec ed]
G igo y Ko zhene ich
g igo y.ko zhene [email protected]
1Ins i u e o Ope a ions Resea ch, Ka ls uhe Ins i u e
o Technology, Ka ls uhe, Ge many
2Cen e o Heal hca e Ope a ions Imp o emen and Resea ch,
Uni e si y o Twen e, Enschede, The Ne he lands
•We show he necessi y o ope a ing oom planning me h-
ods o be es ed on di e en ealis ic se ings since, e.g.,
hospi als o di e en ca e le els exhibi signi ican ly di -
e en case mixes.
•We show bene i s o p ac i ione s o join o se up new
benchma king ini ia i es.
1 In oduc ion
The ope a ing oom (OR) plays a c ucial ole in a hospi al’s
ope a ions since, o mos hospi als, a signi ican ac ion o
ea ed pa ien s and gene a ed e enues a e associa ed wi h
su gical se ices [49]. Because o his and because an OR
is ypically a highly complex sys em wi h many di e en
s akeholde s, expensi e esou ces, ime-sensi i e p ocesses,
and an inhe en ly high le el o unce ain y, op imizing he
e iciency o OR ope a ions h ough adequa e planning is
c ucial.
Resea ch on ope a ing oom planning in Ope a ions
Resea ch is popula and ex ensi e [14,16,31,35,38,41,70,
75,94,99]. To es modeling and solu ion app oaches, inpu
da a is needed, which is he ocus o ou wo k. We belie e
123
A eady- o-use su gical p ocess da a se 329
ha o compa e di e en models and solu ion echniques,
hey should be es ed on di e en da a se s ep esen ing di -
e en OR se ings. He e, so-called benchma king se s can
be used [53]. Benchma king se s ep esen collec ions o
ins ances o pa icula (op imiza ion) p oblems [47]. They
can be based on ic ional (i.e., gene a ed) o eal-wo ld da a
[53]. Since he esea ch on OR planning is implemen a ion-
o ien ed [14], he eal-wo ld da a app oach is mo e desi able.
To gene a e such benchma k se s, eal da a should be col-
lec ed sys ema ically and in a s anda dized manne . The la e
aspec is c ucial o enable compa isons ac oss o ganiza ions
and agg ega ion o mul iple da a sou ces i desi ed.
Howe e , da a collec ion cos s ega ding echnical, o ga-
niza ional, and inancial esou ces a e high [55], while he
pu poses aside om manda o y legal compliance migh no
always be appa en o he decision-make s. Consequen ly,
eal-wo ld da a o he esea ch on OR planning is s ill
sca ce. I eal da a se s a e used, hey a e o en small o
low-dimensional. Typically, he da a om only one hos-
pi al is used [38]. Thus, only his hospi al’s speci ic OR
con ex ega ding o ganiza ion, esou ces, su gical po o-
lio, and p ocedu es is being in es iga ed. In he ace o he
jus -desc ibed sca ci y o eal-wo ld OR da a, i is ema k-
able ha he e is an OR benchma king ini ia i e in he case o
Ge man-speaking coun ies. This ini ia i e has been a ound
o almos 15 yea s. O e 320 1Ge man, Aus ian, and Swiss
clinics eco d and submi hei su gical da a in a s anda dized
way. The da abase o he benchma king p og am con ains
millions o su gical eco ds [9]. Each da a poin ep esen s
a pe o med su ge y and includes da a on di e en su ge y-
ela ed pa ame e s.
Fo he pa icipa ing hospi als, he p ima y pu poses o
he benchma king ini ia i e a e o compa e hei OR pe -
o mance ega ding pa icula KPIs such as OR u iliza ion
among each o he and o e alua e he de elopmen o one’s
pe o mance o e ime [9]. Howe e , we a gue ha he da a
collec ed o benchma king pu poses can also be used o
scien i ic pu poses and esea ch on OR planning.
We ind he da a sui able o s udies on OR planning
o mul iple easons. In a nu shell, he da a shows high
le els o quan i y and quali y, s anda diza ion, and mul i-
dimensionali y. Mul iple p ocess ime s amps a e eco ded
pe su ge y, which enables de ailed modeling o he su gical
p ocess, i.e., by b eaking a su ge y down in o se e al p ocess
s eps. Fo ou pu pose, by “su gical p ocess da a”, we deno e
he da a on su gical p ocess s eps du a ions and conside he
en i e pe iope a i e p ocess as he scope o his de ini ion.
Wea gue ha hesu gicalp ocessda a om heORbench-
ma king ini ia i e o Ge man hospi als especially has he
1As o 2022.
po en ial o de ailed modeling app oaches o he sho - e m
(“ope a ional”) [37] OR planning, i.e., su ge y scheduling,
in pa icula . Howe e , i can also be used o s udies on
OR p ocess design. Rega ding he in es iga ion app oach,
he highly de ailed da a seems mos sui able o simula i e
app oaches and Job-Shop-like models. We no e ha he da a
can be agg ega ed o a lowe le el o de ail o be used as inpu
o low-de ailed ypes o planning models as well.
This s udy aimed o p ocess a da a se om he OR bench-
ma king ini ia i e o Ge man-speaking coun ies o esea ch
on OR planning o he i s ime and o make i eady o el-
low esea che s o use. Fo his, we used he benchma king
da a om 2019 and de i ed di e en OR se ings based on
pa ame e s such as hospi al le el o ca e (LOC) o su gical
special y. Fo each se ing, we ha e calcula ed dis ibu ions
o su gical p ocess du a ions and case mixes o su gical p o-
cedu es, ep esen ing he su gical po olio o he espec i e
OR se ing. One pa icula ocus o ou s udy was o model
a su ge y, no in i s en i e y, bu o dis inguish se e al p o-
cess s eps and o iew hem sepa a ely so ha he da a could
be used in de ailed model app oaches, as men ioned p e i-
ously. Conc e e benchma k se s and p oblem ins ances can
be gene a ed om ou collec ion o su gical case mixes and
p ocess du a ion dis ibu ions. We discuss in de ail how his
could be app oached and sugges se e al OR planning p ob-
lemsand in es iga ionapp oaches o which suchbenchma k
se s could be use ul. The collec ion o case mixes and p ocess
du a ion dis ibu ions can be accessed eely online [48].
One pu pose o ou s udy is o jus i y he p ac ical ele-
ance o he sys ema ic collec ion o su gical p ocess da a
in he con ex o p ospec i e OR planning and o encou -
age hospi als and OR manage s o e-e alua e hei cu en
da a collec ion p ac ices. Joining many ellow esea che s,
we wan o d aw he p ac i ione s’ a en ion o he po en-
ial o (da a-based) OR planning me hods om he ield o
Ope a ions Resea ch.
The pape is o ganized as ollows: In Sec ion 2, we p esen
ela ed li e a u e on su gical p ocess da a and i s use in p ac-
ice and OR planning esea ch, as well as on benchma k se s
and hei sui abili y o es ing di e en modeling and solu-
ionapp oaches.Wealsolis u he exis ingin e na ionalOR
benchma king ini ia i es, which migh ha e a po en ial o
scien i ic s udies simila o he po en ial o he benchma king
p og am we desc ibe he e. In Sec ion 3.1, we p esen he said
ORbenchma kingini ia i eo Ge man-speaking coun iesin
de ail be o e we desc ibe he da a collec ed h oughou he
ini ia i e and how we p ocessed he 2019 da a se and p esen
i in ou da a collec ion in Sec ion 3.2. Sec ion 3.3 discusses
he gene al po en ial and bene i s o he benchma king da a
and speci ically o ou collec ion o su gical case mixes and
p ocess du a ion dis ibu ions o di e en OR se ings. We
123
330 G. Ko zhene ich e al.
inish wi h a de ailed discussion on he limi a ions o he
benchma king da a and ou app oach and sugges ways o
add ess hose issues. Sec ion 4p esen s a concluding sum-
ma y o ou wo k and an ou look o u u e esea ch.
2 Li e a u e e iew
2.1 Su gical p ocess da a in OR p ac ice
In OR ope a ions, collec ing speci ic su gical p ocess da a
can be manda o y o hospi als o quali y assu ance and
accoun ing easons, based on p e alen egula ions [55].
While da a collec ion s anda ds migh no always be manda-
o y, i is essen ial o ensu e consis en documen a ion o e
ime and alid benchma king [7,11]. The da a on p ocess
du a ionsisusually ou inelycollec eddu ing su ge yas ime
s amps o pa icula p ocess miles ones, e.g., OR en y o
incision [7,11]. The da a on su ge y p ocess du a ions a e
being used in p ac ice o e ospec i e pe o mance anal-
ysis [11] as well as o du a ion o ecas ing in p ospec i e
su ge y planning. The la e ep esen s i s own widely elab-
o a ed esea ch ield in he li e a u e [30]. Typically, speci ic
pa ame e s a e iden i ied as signi ican p edic o s o su ge y
du a ion, e.g., su ge y ype o ope a ing su geon [42,86].
2.2 Su gical p ocess da a in s udies on OR planning
Se e al sys ema ic e iews ouch upon he use o su gical
p ocess da a wi hin he e iewed s udies [14,31,35,38].
Howe e , we could no ind any e iews ha would gi e a
comp ehensi e insigh in o his opic. We p esen a sho
summa y o ou li e a u e indings.
Resea che s ei he use eal-wo ld da a o model su gi-
cal p ocess du a ions o gene a e ic ional p oblem ins ances
[38]. The o me is usually p e e ed o ensu e be e imple-
men abili y o he model o algo i hm [14]. In a model, he
da a is used o wo pu poses: To model he ealized and he
p edic ed p ocess du a ion. Bo h can be done de e minis i-
cally, e.g., by using ac ual eco ded du a ions o he o me
[30] and calcula ing mean alues om his o ical da a o
he la e [49,58]. Typically, howe e , he ealized p ocess
du a ions a e modeled s ochas ically [14] by i ing dis ibu-
ions o he p ocess du a ions om empi ical da a [99]. The
p edic ed p ocess du a ions can be modeled using he pa am-
e e s o he i ed dis ibu ions [36] o , al e na i ely, by using
linea eg ession [49] o o he Machine Lea ning algo i hms
[30].
Su gical p ocess da a is ypically ei he g ouped by spe-
ci ic pa ame e s wi hin a s udy o chosen om an o e all
da a se acco ding o he scope o he s udy. This co e-
sponds wi h he abo e esea ch on po en ial p edic o s o
su gical p ocess du a ions [42]. Fo example, he di e en i-
a ion by hospi al o hospi al ype is mos ly done implici ly,
as many s udies use da a om one hospi al. The same holds
o su gical special y unless se e al su gical depa men s a e
being conside ed simul aneously, e.g., o a Mas e Su ge y
Schedule cons uc ion [57] o join su ge y scheduling [36].
Su ge y o pa ien cha ac e is ics can be used o b eak down
he da a u he . Fo example, su ge y u gency (elec i e s.
non-elec i e) [96] o ype o su gical pa ien (inpa ien s.
ou pa ien ) [97], al hough again - many s udies ocus on one
u gency o case ype and choose he da a om hei o e -
all da a se s acco dingly [34]. Su ge y ype can be used as
a g oupe based on he ac ual su gical p ocedu e(s) [80]o
own classi ica ions [36,44,49,54,67,76]. O he po en ial
g oupe s a e pa ien age o diagnosis [54]. Finally, da a clas-
si ica ion can also be done based on esou ces in ol ed in he
su ge y p ocess, i.e., s a membe s [49,54,62], ope a ing
ooms o medical equipmen [69]. We do no gi e a comp e-
hensi e lis he e. The esea ch on su ge y du a ion p edic ion
can be consul ed o u he possible g ouping pa ame e s.
The le el o de ail in su gical p ocess da a and he co e-
sponding modeled su gical p ocess di e among s udies. We
p opose ha his couldbeano he in e es ingaspec o u u e
sys ema ic e iews. Many s udies iew a su ge y as a whole
and ocus on he in aope a i e phase [36,89]. Al hough i
is no always clea in his case wha pa icula p ocess mile-
s ones de ine he “case ime,” mos o he ime, he wheels-in
o wheels-ou du a ion can be assumed [77]. Many s udies
addi ionally conside p e- and pos ope a i e su ge y phases
[4,33,69] and he co espondingspa ial esou ces such as he
p eope a i e holding uni , pos -anes hesia ca e uni (PACU),
o in ensi e ca e uni [38]. Some s udies model u no e (o
cleaning o OR-se up) ime sepa a ely [4,5,12,30,32,49,
61,67,76]. S udies such as Ba un e al. [5], B own e al.
[12], Holmg en and Pe sson [40], Kougias e al. [49], Messe
e al. [61], Ozen e al. [67] model he ac ual p ocess o a
su ge y wi h h ee main p ocess s eps: P e-incision ( akes
place ei he in he OR, i.e., OR en y o incision [12,49,
67], o in a sepa a e p epa a ion oom [40,61]), incision- o-
closu e and pos -incision (i.e., closu e o OR exi [12,49,
67]). Ba un e al. [5] and Ozen e al. [67] addi ionally model
“su geon u no e .”Thisp ocesss eps a simmedia elya e
closu e and occu s pa allel o pos -incision and OR cleaning.
As Messe e al. [61] a e conce ned wi h inding he op imal
numbe o OR ans e ooms, hey addi ionally model he
inwa d ans e o he pa ien in o he OR a ea be o e p e-
incision and he ou wa d ans e a e pos -incision. La o e
Núñez e al. [51] model he p e-incision phase in mo e de ail
and dis inguish be ween ou di e en p epa a ion o se up
s eps: Pa ien , OR, su geon, and u he esou ces. Riise e al.
[74] ocuse en mo eon he su gical esou cesby conside ing
p ocess s eps such as “ emo al o any supe luous equipmen
om he ope a ing oom” o “ emo al o used equipmen .”
No e ha especially o he in aope a i e phase, i.e., he
123
A eady- o-use su gical p ocess da a se 331
ac ual su gical in e en ion, an ex emely high le el o de ail
in p ocess modeling can heo e ically be achie ed by iden i-
ying indi idual su gical manipula ions [59,64,100]. Such
a high le el o de ail can help es ima e he ( emaining) du a-
ion o a pa icula su ge y [3,100]. Howe e , o he ype
o OR esou ce-planning p oblems we ocus on he e, such a
high le el o p ocess de ail is unnecessa y.
F om ou li e a u e e iew, we ind ha highly de ailed
su gical p ocess da a has been used o simula ion s udies
- ei he o gene a e inpu o su ge y schedule op imiza ion
models [4,67] o o in es iga e he ela ionships be ween
sys em pa ame e s and hei impac on he OR pe o mance
[61]. The e ec o di e en s a is ical me hods o p o-
cess du a ion p edic ion on he OR pe o mance has been
analyzed as well [49]. The da a is also used o de ailed
modeling o pe iope a i e esou ces and hei p ocess s ep-
speci ic alloca ion wi h p ojec -scheduling- ela ed [74]o
low-shop- ela ed app oaches [51]. Wi h de ailed p ocess
da a, o e lapping p ocesses can be modeled, which is, o
example, di ec ly being used by he esea ch on OR p ocess
design [5,12,40]. A high le el o p ocess de ail is gene ally
no necessa y o o he ypical OR planning p oblems on he
s a egico ac icalle el,suchasdimensioningandalloca ing
OR esou ces [37]. De ailed p ocess da a could, howe e , be
used o he s a egic p oblem o layou planning [63], whe e
he ocus lies on he pa hways o he di e en s akeholde s
in he OR. On he ope a ional le el o planning, de ailed p o-
cess modeling enables mo e ealis ic modeling in gene al [5],
and indi idual du a ion modeling o each p ocess s ep, e.g.,
dis ibu ion i ing [67].
As men ioned, in less de ailed su gical p ocess da a, su g-
e ies a e usually conside ed as a whole, i.e., wi h only one
p ocess s ep. Then benchma king se s wi h ealized su ge y
du a ions pe su ge y ype [36] o i ed heo e ical dis-
ibu ions pe ype oge he wi h ixed capaci y alloca ion
decisions and case mixes can be used o su ge y scheduling.
He e, su ge y scheduling on he ope a ional le el is usu-
ally di ided in o ad ance scheduling, i.e., su gical cases a e
assigned o an ope a ing oom on a speci ic day, and alloca-
ion, i.e., sequencing o he su ge ies, po en ially assigning
s a imes. In addi ion, ha da a may be used o eschedul-
ing, e.g., i elec i e su ge ies ha e o be pos poned due o
a i ing eme gencies. Fo example, Jung e al. [44] p esen
op imiza ion models o ad ance and alloca ion scheduling
as well as a escheduling p ocedu e. Dex e and T aub [18]
in es iga es su ge y scheduling heu is ics ia simula ion, and
Landa e al. [50] conside ad ance and alloca ion schedul-
ing wi h s ochas ic su ge y du a ions, also using wai ing lis
da a.
When going o highe le els o planning, less de ailed
da a is usually used. On he ac ical le el, ope a ing oom
capaci y is alloca ed o di e en pa ien g oups, e.g., h ough
block scheduling and ixing a mas e su gical schedule. This
is o en done on he le el o special ies. Fu he , s a ing and
os e ing decisions o ope a ing oom s a a e usually based
on he mas e su gical schedule [8,23]. Vanbe kel e al. [92]
ela e he mas e su gical schedule o he esul ing capaci y
usage o downs eam esou ces such as wa d beds. He e, o
e e y special y, hey assume a dis ibu ion o e he numbe
o su ge ies ha can be pe o med in a su ge y session. Jung
e al. [44] alloca e capaci y o elec i e su ge ies such ha
eme gency pa ien s can also be ea ed. To his end, hey
classi y su ge ies in o sho , medium, and long su ge ies.
Finally, on he s a egic le el o planning, su gical p ocess
da a can suppo se ice design, case mix, and capac-
i y dimensioning decisions. On his le el, models usually
assume de e minis ic alues, such as demand olume o
equi ed capaci y pe pa ien ype [39]. In addi ion, in o ma-
ion is needed on cos s and p o i s o se ing ce ain pa ien
ypes. Fo example, Blake and Ca e [10] p opose a goal p o-
g amming app oach o decide on he case mix and olume
o physicians using de e minis ic alues o needed su ge y
and wa d capaci y pe pa ien ype.
2.3 Benchma k se s
Aswe p esen a newda acollec ion inSec ion 3.2 anddiscuss
i susagepo en ial o ORplanning esea ch,wea ein e es ed
in how su gical p ocess da a like ou s can be made eady
o use by ellow esea che s. We use he wo k by Lee ink
and Hans [53] as guidance o p epa ing he da a so ha
benchma k se s can be de i ed om i . Benchma k se s a e
c ucial o pe o mance compa ison o solu ion app oaches
on di e en p oblem scena ios since no all me hods pe -
o m equally well in all si ua ions [53]. Some s udies p o ide
benchma k se s o gene al low o job shop p oblems [17]o
p esen gene ic p oblem ins ance gene a o s [87]. Mos s ud-
ies on OR planning, such as [50], de ine hei own ins ance
se s. Some make hem publicly a ailable, like Riise e al.
[74], [81].
Lee ink and Hans [53] ocus speci ically on gene a ing
benchma k se s o su ge y scheduling p oblems. They p o-
pose ha a su ge y scheduling ins ance should be de ined
by a su gical case mix and dis ibu ion pa ame e s o each
ype o su ge y in he case mix, including expec ed su ge y
du a ion and a ia ion. No e ha a su ge y in a p oblem
ins ance, as desc ibed by Lee ink and Hans [53], is seen in
i s en i e y, wi hou being di ided in o sepa a e p ocess s eps.
The au ho s sugges an app oach o cha ac e izing he case
mix o a p oblem ins ance and gene a ing se e al su ge y
scheduling ins ances, heo e ical and based on eal-li e da a
om i edi e en Du ch hospi als.LikeRiisee al. [74], hey
make hei benchma k se s publicly a ailable [88]. Se e al
s udies ha e al eady used hese benchma k se s since [38].
123
332 G. Ko zhene ich e al.
Lee ink and Hans [53] conclude hei wo k by sugges ing a
me hod o de e mining he p oximi y o p oblem ins ances
in a pa icula benchma k se and subsequen selec ion o he
leas simila ins ances o ensu e he equi ed di e si y o he
benchma k se .
2.4 Su gical p ocess da a benchma king ini ia i es
We p e iously men ioned ha su gical p ocess da a could be
used o benchma king pu poses. Since we p esen a bench-
ma king p og am es ablished by p o essional associa ions
om Ge many, Aus ia, and Swi ze land in Sec ion 3,we
sho ly lis simila ini ia i es om o he coun ies.
Su p isingly, we did no ind many examples o na ional
OR benchma king ini ia i es. We s a by naming wo u he
Ge manini ia i essimila o heone we ocusonin hiss udy.
One is he benchma king ini ia i e by K ankenhauszweck-
e band Rheinland wi h 87 pa icipa ing hospi als in 2020
[46]. The o he is adminis a ed by BIno is GmbH and JR
Consul ing oHG and claims i s unique app oach by e alua -
ing o ganiza ional aspec s o an OR, addi ionally o ypical
p ocess KPIs [43]. Fo English-speaking coun ies, we ind
e idence ha in 2011, 471 hospi als and ambula o y su ge y
cen e s om he USA, Canada, Saudi A abia, Aus alia, and
New Zealand pa icipa ed in he so-called “OR Benchma ks
Collabo a i e,” un by McKesson En e p ise In elligence,
USA [25]. Boggs e al. [11] upda e he P ocedu al Times
Glossa y (PTG) o he US Associa ion o Anes hesia Clin-
ical Di ec o s and no e ha he PTG has al eady acili a ed
benchma king ini ia i es. Un o una ely, he au ho s do no
name any examples. Simila ly o he PTG, ope a ing he-
a e e iciency guidelines exis in Aus alia [1,82]. We ound
e idence ha he Na ional Heal h Se ice (NHS) England,
speci ically he NHS Benchma king Ne wo k, epo s annual
benchma king esul s in i s “Ope a ing Thea es P ojec ”
[66], wi h 69 hospi als pa icipa ing in 2018 [65]. The esul s
include insigh s on OR pe o mance indica o s such as u i-
liza ion o u na ound ime [65,66].
We ound only one na ional benchma king ini ia i e om
a non-Ge man-speaking coun y ha we conside simila o
heone we ocuson in his pape : Thebenchma king p og am
o he uni e si y hospi als in he Ne he lands, es ablished in
2005. The su gical p ocess da a o he se en pa icipa ing
clinics a e p ocessed and analyzed cen ally. The hospi als
egula ly ecei e insigh s on he e iciency and p o i abili y
o hei ORs compa ed o ellow benchma king pa icipan s.
The pa icipa ing clinics a e encou aged o exchange bes
p ac ices wi h each o he . The collec ed da a can be p o ided
in anonymized o m o scien i ic s udies. The le el o de ail
in he da a is high, wi h se e al ime s amps co esponding o
he su gical and anes he ic p ocedu es collec ed pe su ge y
[90,91].
3 The OR benchma king p og am su gical
p ocess da a and i s po en ial o OR
planning esea ch
3.1 The OR benchma king p og am
o Ge man-speaking coun ies
3.1.1 The Ge man Pe iope a i e P ocedu al Time Glossa y
In 2008, he i s e sion o he “The Ge man Pe iope a i e
P ocedu al Time Glossa y” (GPPTG) was published, ol-
lowing he eme ging demand o a s anda dized, KPI-based
OR managemen and ex e nal benchma king among Ge man
hospi als[9]. TheGlossa y was hep oduc o a join e o by
he Ge man p o essional associa ions o anes he is s (BDA),
su geons (BDC), and OR manage s (VOPM). The GPPTG
has been e ised and upda ed wice since - in 2016 and 2020.
In he 2020 e sion, he Aus ian and Swiss associa ions o
OR manage s (VOPMÖ and SFOPM, espec i ely) became
in ol ed as well, ex ending he alidi y o he GPPTG o all
h ee Ge man-speaking coun ies. In i s mos ecen e sion,
he Glossa y con ains 41 de ined pe iope a i e p ocess ime
poin s, ca ego ized in o subca ego ies pa ien logis ics, OR
logis ics, anes hesia, and ope a ion. Su gical p ocess s eps
based on hese ime poin s and ypical KPIs conce ning he
ORpe o mancea e also de ined.Howe e , he imepoin s o
he p ocess s eps do no sugges a “s anda d” su ge y p ocess
bu a he encompass ypical miles ones o a gene ic su gical
p ocess. The en i e pe iope a i e p ocess is co e ed, om he
pa ien being called o he pa ien being discha ged om he
PACU. Howe e , he GPPTG ocuses on he pa ien ’s pa h
h ough he OR, so o he OR- ela ed asks, such as documen-
a ion o planning, a e no included [7].
3.1.2 Benchma king p og am
In connec ion wi h he ini ial publica ion o he GPPTG, he
a o emen ioned benchma king p og am o su gical p ocess
da a was es ablished in 2009. F om he ou se , i s cen al
pu pose has been o p o ide pa icipa ing hospi als wi h an
oppo uni y o compa e OR pe o mance among each o he
and, wi h his, o e alua e one’s po en ial o imp o emen .
The echnical implemen a ion is ca ied ou by a neu al pa y
company (digmed GmbH, Hambu g, Ge many). A pa ici-
pa inghospi al ypicallysubmi s all i s ou inely eco dedOR
p ocess da a mon hly. The da a collec ion i sel mus ollow
he GPPTG. Pa icipa ion in he benchma king is possible
by submi ing a leas wo ime s amps pe su ge y: Inci-
sion and closu e. Addi ional equi ed in o ma ion o each
su ge y in ol es he da e, he su gical depa men , he ope -
a ing oom, and he unique (anonymized) iden i ica ion o he
ope a ed pa ien [9]. In p inciple, pa icipa ion in he bench-
ma king p og am is open o any Ge man, Swiss, o Aus ian
123
A eady- o-use su gical p ocess da a se 333
hospi al. Howe e , he benchma king esul s a e p o ided
o he benchma king pa icipan s only, excep o scien i ic
s udies. Anonymized da a can be p o ided o he la e [9],
and he e a e al eady s udies ha use he benchma king da a
o esea ch on OR pe o mance [21].
The numbe o hospi als pa icipa ing in he p og am has
g own om 20 hospi als in 2009 [9] o o e 320 Ge man,
Aus ian, and Swiss clinics oday [20]. Among he hospi als,
all le els o ca e (LOC) a e ep esen ed [9,20].2
3.2 Su gical p ocess da a om he OR
benchma king ini ia i e
3.2.1 The 2019 da a se
A da a se om he p e iously desc ibed benchma king p o-
g am was kindly p o ided o us by digmed GmbH. We use
his da a o de i e p ocess du a ions and case mix dis ibu-
ions. The da a se includes all su gical da a o 2019 and all
pa icipa ing Ge man hospi als (Aus ian and Swiss hospi als
we e no included). The e ec o he COVID-19 pandemic on
he OR ope a ions in Ge man hospi als and hus on he co e-
spondingda ainyea ss a ing2020isnon-neglec able,so he
2019 benchma king da a ep esen s he la es non-COVID-
a ec ed si ua ion o Ge man ORs. In he da a se , which
we call he 2019 da a se , 212 hospi als a e ep esen ed in
o al, which accoun s o a ound 11% o all Ge man hospi-
als [83]. The 2019 da a se consis s o 2,035,126 da a poin s,
i.e., eco ded su ge ies.3Fo each su ge y, he unique hos-
pi al ID, he hospi al’s ede al s a e, he hospi al LOC, he
su gical special y,4 he su ge y da e, he OR, and he unique
IDo he pa ien ’shospi al s ay a e eco ded. Fu he su ge y-
speci ic pa ame e s a e op ional and no always eco ded o
all da a poin s o by all hospi als. Those pa ame e s include
hemainOPS
5code o he ope a ion, he anes hesia ype
2Hospi al LOC is, in his case, he Ge man classi ica ion o acu e ca e
clinics based on clinic size (measu ed in beds) and specializa ion o
he medical o e ing. No e ha he hospi al LOC classi ica ion in he
benchma king p og am is pa ly p o isional since he e is no s anda d-
ized classi ica ion ac oss all Ge man ede al s a es so a [95]. The e ms
used in he benchma king p og am also co espond only app oxima ely
o he acco ding classi ica ions in Aus ia [29], and Swi ze land [13].
3We use he e m su ge y synonymously o wha in he GPPTG is
de ined as ope a ion. The la e can consis o one o mo e p ocedu es.
One o mo e su ge ies make up a session [7].
4He e we use he codes published by he Ge man Hospi al Fede a ion
(“Deu sche K ankenhausgesellscha ”) and he Na ional Associa ion o
S a u o yHeal hInsu anceFunds(“GKVSpi zen e band”).Thecoding
app oach can di e ac oss ede al s a es [71].
5The main p ocedu e o he su ge y wi h espec o he “Ope a ionen-
und P ozedu enschlüssel,” i.e., Ope a ion and P ocedu e Code [72],
which is he Ge man modi ica ion o he In e na ional Classi ica ion o
P ocedu es in Medicine [24].
Table 1 GPPTG imes s amps included in he 2019 da a se
Time s amp code Time s ampa
P2 Pa ien A i al a OR sui e
P5 Pa ien In OR
P7 Pa ien Ou o OR
P8c S a PACU
P10 End OR Cleaning
A6 S a Anes hesia
A7 Anes hesia Ready
A9 End Anes hesia
O8 Incision
O10 Closu e
O11 End Follow-up Su gical Measu es
aThe English e ms a e aken om GPPTG 2020, al hough in 2019 he
p e ious 2016 e sion was s ill alid. The la e , howe e , had no been
ansla ed in o English
(local o no local, i.e., gene al anes hesia), he ype o su gi-
cal pa ien (inpa ien o ou pa ien ), he u gency (elec i e o
co esponding o a pa icula le el o eme gency, ollowing
he GPPTG classi ica ion [7]), he main ope a ing hou s o
he co esponding su gical special y (K18a [7]) and he size
o he OR block capaci y assigned o he su gical special y
in he pa icula OR and on he pa icula da e (K18 [7]).
In Table 1, all GPPTG imes s amps included in he 2019
da a se a e lis ed.6Table 2includes all p ocess imes, which
can be calcula ed using hese ime s amps as de ined by he
GPPTG.
digmed GmbH conduc s da a plausibili y checks o ensu e
high da a quali y [9]. The la e is equi ed o benchma k-
ing analyses and scien i ic s udies [45,78,79]. I should be
no ed ha hospi als ha join he benchma king p og am end
o imp o e he quali y o he eco ded su gical p ocess da a
(some imes ema kably) o e ime [9]. In ou 2019 da a se ,
da a poin s a e ma ked i hey ha e passed he plausibil-
i y checks by digmed GmbH. These plausible da a poin s
accoun o mo e han 98% o he da a se . We addi ion-
ally check how well he op ional su ge y pa ame e s and
he p ocedu al ime s amps a e documen ed. In Table 3 o
each su ge y pa ame e , he pe cen age o da a poin s wi h a
de ini e en y, i.e., a eco ded alue excluding he unknown
alues, in he o al da a se a e lis ed. In Table 4, o e e y
ime s amp, we lis he pe cen age o he en i e 2019 da a
se ha has he ime s amp eco ded and is a he same ime
ma ked plausible by digmed GmbH.
6digmed GmbH had da a on o he ime s amps a ailable; howe e ,
hose we e submi ed by only a ew hospi als and esul ed in oo ew
da a poin s o ou pu pose.
123
334 G. Ko zhene ich e al.
Table 2 P ocess imes based on he a ailable GPPTG ime s amps in
he 2019 da a se
P ocess s amp code P ocess s ampb
K2 Anes hesia Induc ion Time (A6 o A7)
K3 Anes hesia Eme gence Time (O11 o A9)
K7 Su gical Lead-in (A7 o O8, o P5 o O8, i
P5 a e A7)
K8 Incision- o-Closu e Time (O8 o O10)
K9 Su gical Lead-ou (O10 o O11)
K10 Pe iope a i e Time (A7 o O11, o P5 o
O11 i no anes hesia used)
K13 Ne Anes hesia Time (A6 o A9)
K15b Tu no e Time Anes hesia (O11 o A7 o
he ollowing case)
K16 Closu e- o-Incision Time (O10 o O8 o he
ollowing session)
K17 Column Time (P5 o P7)
K17a Room Occupied Time (P5 o P10)
bThe English e ms a e aken om GPPTG 2020, al hough in 2019 he
p e ious 2016 e sion was s ill alid. The la e , howe e , had no been
ansla ed in o English
3.2.2 Da a p ocessing
Following Lee ink and Hans [53], one goal is o de e -
mine su gical case mixes om he da a o di e en OR
se ings. We de ine he la e using speci ic pa ame e s and
g oup he aw da a acco dingly du ing da a p ocessing. We
use he hospi al LOC and he su gical special y as se ing
o g ouping pa ame e s ollowing he app oach desc ibed in
Sec ion 2. This di e en ia ion is easonable since hospi als o
di e en LOCs and su gical special ies ypically ha e di e -
ing su ge y po olios ega ding he p ocedu es pe o med.
The o ganiza ion, including he p ocesses and he esou ces,
migh also di e . We use he (main) OPS code o ep esen
he su ge y ype. Based on an addi ional analysis du ing da a
Table 3 Pe cen age o da a poin s in he 2019 da a se wi h a de ini e
en y pe su ge y pa ame e
Su ge y pa ame e cPe cen age o da a poin s wi h a
de ini e alue in he 2019 da a se
(Main) OPS code 89%
Anes hesia ype 60%
Type o su gical pa ien 90%
U gency 83%
Main ope a ing hou s o
he su gical special y
92%
cWe do no conside he block capaci y he e since a missing alue, in
his case, does no necessa ily indica e missing da a bu could mean ha
he su gical special y didn’ ha e any capaci y alloca ed in his OR on
his da e
Table 4 Pe cen age o da a poin s om he 2019 da a se wi h eco ded
and plausible ime s amps
Time s amp Pe cen age o da a poin s wi h ime s amp
eco ded and plausibili y check passed in
he 2019 da a se
P2 73%
P5 30%
P7 27%
P8c 6%
P10 3%
A6 82%
A7 85%
A9 80%
O8 98% (all plausible da a poin s)
O10 98% (all plausible da a poin s)
O11 91%
p ocessing, we decided o use he ype o su gical pa ien
as ano he se ing pa ame e . Table 5shows he alues we
choose om he da a o each pa ame e . The u gency does
no seem o ha e a signi ican addi ional e ec on he p ocess
du a ions.
Rega ding he pa ame e anes hesia ype, we only con-
side su ge ies no explici ly ma ked as ca ied ou in local
anes hesia since he g oup ep esen s less han 1% o ou
inal da a selec ion. See Appendix A o mo e de ails on ou
da a selec ion p ocedu e. To de e mine he case mix o each
combina ion o hospi al LOC, su gical special y, and ype o
su gical pa ien , we de e mine he OPS codes ep esen ed in
he co esponding da a selec ion and hei ela i e equency
in he conside ed class.
Fo each su ge y ype in a pa icula case mix, we i heo-
e ical dis ibu ions ( wo-pa ame e logno mal, gamma, and
Weibull) o p ocess du a ions based on he his o ical da a,
simila o Lee ink and Hans [53]. O he han Lee ink and
Hans [53], we aim o a mo e de ailed modeling o he su -
gical p ocess han conside ing a su ge y in i s en i e y. We
choose he p ocess-o ien ed pe spec i e based on a ailable
ime s amps, ocusing on he main pe iope a i e ac i i ies,
i.e., anes hesia and su gical p ocedu es. We de e mine he
Table 5 Chosen alues pe pa ame e o p ocess he 2019 da a se
Hospi al LOC Su gical special y Type o su -
gical pa ien
Basic and Regula Ca e Gene al Su ge y Inpa ien
Specialized Ca e T auma Su ge y Ou pa ien
Uni e si y Clinics O ola yngology
Maximum Ca e, excl.
Uni e si y Clinics
Gynecology
and Obs e ics
123
A eady- o-use su gical p ocess da a se 335
Fig. 1 P ocess s eps o a gene ic su ge y acco ding o he p ocess s amps de ined in Table 2
mos de ailed and consecu i e p ocess modeling possible
and choose K2, K3, K7, K8, and K9 as ou main p ocess
s eps o imes. We assume he p ocess s eps o ep esen a
gene ic su ge y as shown in Fig. 1. We no e he e ha his
assumed sequence o p ocess s eps and ime s amps is jus
one o many possibili ies o how he su gical p ocess in a
Ge man hospi al could be modeled o implemen ed in eal-
i y. As indica ed in Table 2, men ioned in 3.1.1 and depic ed
in de ail in Baue e al. [7] he e a e di e en possibili ies
o he p ocess design o OR logis ics, conce ning o exam-
ple he anes hesia p ocedu e o he pa ien logis ics, which
can esul in di e en p ocess imes de ini ions in e ms o
su gical ime s amps. We addi ionally i dis ibu ions o he
OR cleaning ime (which we de ine as A9 o P10), so his
p ocess s ep could be modeled sepa a ely i desi ed. Finally,
we i dis ibu ions o he closu e- o-incision ime (K16),
al hough no ecommended bu s ill used by some Ge man
hospi als p oxy o he u no e ime [77]. See Appendix B
o mo e de ails on ou dis ibu ion i ing me hod. Fo ou
inal collec ion o he p ocess du a ion dis ibu ions, we cal-
cula e he expec ed alue and a iance besides he es ima ed
dis ibu ion pa ame e s o each dis ibu ion.
3.2.3 Ou collec ion o p ocess ime dis ibu ions
and su gical case mixes
The main ou pu o ou p e iously desc ibed analysis and
p ocessing o he 2019 da a se is ou collec ion o p ocess
ime dis ibu ions and su gical case mixes. The collec ion o
su gical case mixes is ep esen ed by a sp eadshee wi h ou
pa ame e columns: Hospi al LOC, su gical special y, ype
o su gical pa ien , and (main) OPS code. The ep esen ed
alues o he i s h ee pa ame e s a e lis ed in Table 5.
Mo eo e , 633 unique OPS codes a e ep esen ed in ou case
mix collec ion. The e a e 1,685 unique combina ions o he
ou pa ame e s in ou inal case mix collec ion. The case
mix sp eadshee includes he co esponding class size o
each unique combina ion. I is exp essed by he numbe o
obse a ions, i.e., da a poin s o unique su ge ies, om ou
main da a se (see Appendix A). See Fig. 2 o an exce p o
he case mix sp eadshee .
Table 6shows o each unique combina ion o hospi al
LOC, su gical special y, and ype o pa ien ha is ep e-
sen ed in ou case mix collec ion, he numbe o included
OPS codes ha co espond wi h ha pa icula pa ame e
combina ion, he o al numbe o obse a ions summed up
o e all hese OPS codes, he a e age class size o he indi-
idual OPS codes as well as he s anda d de ia ion o he
class size. I can be obse ed ha all combina ions excep
o he combina ion o Uni e si y Clinics, Gene al Su ge y,
and Ou pa ien a e ep esen ed in ou case mix collec ion.
The inpa ien combina ions ypically include a much la ge
o al numbe o obse a ions and OPS codes han hei ou -
pa ien coun e pa s. The numbe o included OPS codes pe
combina ion anges be ween 4 (Uni e si y Clinics, T auma
Su ge y, Ou pa ien ) and 182 (Specialized Ca e, Gene al
Su ge y, Inpa ien ). Table 7shows he i e la ges OPS
codes ep esen ed in he case mix collec ion as measu ed
Fig. 2 An exce p om ou inal case mix collec ion
123
342 G. Ko zhene ich e al.
agg ega e ac oss mul iple hospi als, anes hesia p ocedu es,
u gency le els, su geons, and o he esou ces and do his o
a yea . This limi s he ep esen a i eness o ou de i ed case
mixes and dis ibu ions. To calcula e he la e , we agg ega e
he benchma king da a ac oss hospi als. Hence, ou esul ing
collec ion is less sui able o a de ailed analysis o one pa ic-
ula hospi al and i s indi idual OR ope a ions. I s po en ial
lies hus p ima ily wi h a mo e gene ic scope o esea ch,
al hough he p ocess du a ion dis ibu ions could be used as
a p oxy i a hospi al’s da a is sca ce, as men ioned p e iously.
Du ing ou da a selec ion p ocess, we had o mee se -
e al mo e o less a bi a y assump ions, e.g., which p ocess
du a ions we conside implausible. As a esul o he da a
selec ion, we excluded as much as 90% o he o iginal 2019
da a se o ob ain he inal main da a se ha we used o
de i e case mixes and du a ion dis ibu ions. This na u ally
con ibu es u he o he ep esen a i eness issue. The la ge
pe cen age is mainly due o h ee majo goals o ou da a
selec ion p ocess: (1) High da a plausibili y ( alid pa ame e
alues); (2) exclusion o i egula su ge y se ings (ope a -
ing ou side egula opening hou s, o e lapping p ocess s eps,
local anes hesia p ocedu es); and (3) high le el o de ail
(numbe o ime s amps and g ouping pa ame e s). Wi h he
la e , we also wan ed o ensu e a su icien class size o each
unique combina ion o g ouping pa ame e s in he main da a
se . Thus, we only conside ed he ou la ges hospi al LOCs
and he ou la ges su gical special ies om he 2019 da a
se as lis ed in Table 5and emo ed all combina ions wi h
class sizes o less han 30 da a poin s in he inal da a selec-
ion. Resea che s who wish o use ou da a collec ion o
hei s udies should be awa e o he ac ha i ep esen s
hese pa icula OR se ings only. See Appendix A o mo e
de ails on ou da a selec ion p ocess.
When de i ing p ocess du a ion dis ibu ions, we encoun-
e ed goodness-o - i issues ha we had o deal wi h in e e y
ou h case (see Appendix B). We conclude ha he du a ion
dis ibu ions we ha e i ed canno depic e e y aspec o he
ac ual su gical p ocess da a p ecisely each ime. I is possible
ha , in some cases, heo e ical dis ibu ions o he han hose
weconside ed(gamma,logno mal,Weibull)migh ha ebeen
he be e choice. I is also possible ha in some cases,
an unimodal dis ibu ion was no he igh app oach in he
i s place, whe e, o example, a bimodal dis ibu ion would
ep esen he empi ical da a mo e accu a ely [56]. Figu e 8
shows his in he con ex o uni e si y clinics, o ola yngol-
ogy, inpa ien s, OPS code 5-059.c7, and incision- o-closu e
ime.
4 Conclusion
To conclude, we summa ize he esul s o ou wo k: We
ha e p esen ed he OR benchma king ini ia i e o Ge man-
Fig. 8 Empi ical bimodal and app oxima ed unimodal dis ibu ion o
he incision- o-closu e ime o he OPS code 5-059.c7 (o ola yngology
inpa ien s in a uni e si y clinic)
speaking coun ies in he con ex o esea ch on OR planning
o he i s ime. We elabo a ed in de ail on he p ope ies
o he su gical p ocess da a collec ed in he benchma k-
ing p og am and i s po en ial o OR planning esea ch.
Fu he , we made he p ocessed da a eely a ailable, so el-
low esea che s could use i o es modeling and solu ion
app oaches o di e en OR planning p oblems. Co espond-
ing o ou da a selec ion p ocess, di e en OR se ings
de e mined by he hospi al LOC, he su gical special y, and
he ype o su gical pa ien can be in es iga ed using he da a
collec ion o su gical case mixes and p ocess du a ion dis i-
bu ions we p o ide. Since we b eak down he pe iope a i e
su gical p ocess in se e al sepa a e s eps in ou da a, i is o
pa icula ele ance o highly de ailed app oaches such as
simula ion o Job-Shop-like models, especially on he ope -
a ional le el o planning. Howe e , when agg ega ing and
ex ending ou p o ided da a wi h addi ional in o ma ion, i
can be used o OR planning p oblems on all planning le -
els. Finally, we ha e discussed he bene i s and limi a ions
o he benchma king p og am, he collec ed su gical p ocess
da a, and ou da a p ocessing app oach and i s esul s. Wi h
i s as da a collec ion, we a gue ha he benchma king ini-
ia i e poses a unique oppo uni y o scien i ic esea ch on
OR ope a ions.
We sugges se e al di ec ions o u he s udies and o
applying ou esul s in he ollowing. Fo he esea che s
who wan o use he da a collec ion o su gical case mixes
and p ocess du a ion dis ibu ions we p o ide o gene a e
benchma k se s o p oblem ins ances, we ecommend using
he me hods desc ibed in Lee ink and Hans [53].
123
A eady- o-use su gical p ocess da a se 343
Following ou sugges ion o use he da a o he OR bench-
ma king ini ia i e o Ge man-speaking coun ies o he
esea ch on OR planning, we encou age ellow esea che s
o con inue o wo k on and de elop planning models and
me hods ha a e highly de ailed in e ms o using high-
dimensional inpu da a and modeling he pe iope a i e su gi-
cal p ocess wi h se e al sepa a e p ocess s eps in pa icula .
The e a e s ill ew such app oaches oday, which migh be
because he e was no much de ailed eal-wo ld da a a ail-
able p e iously.
We u he belie e ha a sys ema ic e iew a icle on
using eal-wo ld su gical p ocess da a and he co esponding
modeled su gical p ocesses would signi ican ly con ibu e o
Ope a ions Resea ch in OR planning.
We wan o sugges se e al u he possibili ies o p o-
cessing su gical p ocess da a om he OR benchma king
ini ia i e ha we ha e p esen ed he e. Fi s , he o iginal 2019
da a se ha we used could be p ocessed and p epa ed di e -
en ly, as we did he e, in a mo e sui able way o a pa icula
esea ch pu pose. This could, o example, be done by imple-
men ing a di e en da a selec ion app oach. Al e na i ely o
ou p ocedu e o i ing he du a ion dis ibu ions o sepa-
a e su gical p ocess s eps, he su gical case ime, i.e., he
du a ion o a su ge y as a whole, could be in he ocus. Since
dis ibu ion i ing migh be di icul o po en ially mul i-
modal dis ibu ions, we ecommend di ec ly wo king wi h
benchma k se s.
Ano he esea ch pa h would be o use da a mo e ecen
han 2019, o example, o in es iga e he e ec o he
COVID-19 pandemic on OR ope a ions. This would also be
a possibili y o include mo e da a pa ame e s in he analysis.
A deepe di e in o he da a analysis migh also be o in e es :
An app oach di e en om ou s could be employed o he
dis ibu ion i ing o p ocess du a ions. In some cases, one
could es whe he heo e ical dis ibu ions o he han log-
no mal, gamma, o Weibull migh i be e . In o he cases,
mul imodal dis ibu ions migh be a p omising app oach, as
Sec ion 3.3.2 men ions. An ensuing esea ch ques ion would
be whe he mul imodal du a ion dis ibu ions equi e new,
speci ic planning app oaches since exis ing planning models
and me hods ypically deal wi h unimodula dis ibu ions.
Finally, an ex ensi e analysis o di e en da a pa e ns in
he benchma king da a, e.g., dependencies and co ela ions
be ween indi idual p ocess imes, using elabo a e da a anal-
ysis me hods is mos likely ogene a e new aluable insigh s.
The benchma king da a could become e en mo e aluable
o OR decision-making i addi ional a ibu es we e col-
lec ed, e.g., in o ma ion on he se up, such as he unde lying
mas e su gical schedule and applied scheduling p ocedu es,
he usage o (downs eam) esou ces and s a , and he ac ual
demand o su ge y, including wai ing lis s.
As o he OR benchma king ini ia i e ha we ha e p e-
sen ed in his pape , we hope o inc ease awa eness o he
pa icula esea ch ield o OR planning and he inhe en
po en ial o he benchma king da a in his ega d. Any ini ia-
i e acili a ing u u e scien i ic endea o s in his ield, such
as au oma ed da a p ocessing (e.g., du a ion dis ibu ion i -
ing) as pa o he egula benchma king ope a ions, would
be e y welcome.
A he e y end, we wan o use he inal oppo uni y o
add ess p o essional associa ions o su geons and anes hesi-
ologis s, OR manage s, and Ope a ions Resea ch scien is s in
he ield o OR planning om o he coun ies and encou age
hem o pu sue OR benchma king ini ia i es and le e age he
po en ial o exis ing p ojec s in a simila way ha we did in
his s udy. Speci ically, we mean p ocessing su gical p ocess
da a and p o iding he esul s wi h ee access as we did.
We hink ha an in e na ional da abase o su gical p ocess
da a benchma k se s om di e en coun ies could be a e y
p omising endea o o he en i e esea ch ield and, hus, o
OR ope a ions a ound he globe.
Appendix A Da a selec ion
We ha e o exclude he da a o one pa icula hospi al om
he o iginal benchma king da a se due o inco ec o ma -
ing o he p ocessing imes. We also exclude all su ge ies
wi hincisions akingplaceou sideo 2019.We call he esul -
ing da a se he 2019 da a se as desc ibed in Sec ion 3.2.
We ocus only on he egula OR ope a ions. The e o e, we
exclude all eco ds o su ge ies ha ook place on weekends
o public holidays, as well as su ge ies, ca ied ou ou side
he main ope a ing hou s o he espec i e su gical depa -
men o on days o which he depa men did no ha e any
su gical capaci y in he espec i e OR assigned.
Mo eo e , we only keep he su gical eco ds ha a e
ma ked plausible by digmed GmbH, ha e he main OPS code
eco ded, he main ime s amps we a e in e es ed in (A6, A7,
A9, O8, O10, O11) eco ded, and he co esponding p o-
cess ime (K2, K3, K7, K8, K9) s ic ly g ea e ze o. O he
han Messe [60], we do no allow any p ocess ime o be
ze o. We keep elec i e as well as u gen su ge ies. Howe e ,
we simpli y he classi ica ion by g ouping all su ge ies wi h
an eme gency le el assigned in o one eme gency g oup. We
addi ionally iden i y consecu i e su ge ies in he da a se ,
i.e., su ge ies ha ook place in he same hospi al, OR, and
da e. We de e mine he closu e- o-incision ime (K16) o
each pai o consecu i e su ge ies, as well as whe he he
anes hesia s a (A6) o he la e su ge y ook place be o e
he anes hesia end (A9) o he p e ious su ge y. We ma k
such o e lapping su ge ies and exclude hem om u he
123
344 G. Ko zhene ich e al.
conside a ion since p ocess du a ions o o e lapping su g-
e ies co espond o a di e en p ocess o ganiza ion and di e
signi ican ly compa ed o he s ic ly consecu i e su ge ies,
acco ding o Schus e e al. [79]. We assume ha all o he
su ge ies in ou da a se we e non-o e lapping.
We choose he ou la ges hospi al LOCs in ou da a se
and he ou la ges su gical special ies as lis ed in Table 5.
We con inue by ounding all p ocess imes o he nea -
es minu e and ca y ou ou seconda y plausibili y checks:
We il e ou all su ge y eco ds i ei he p ocess imes
anes hesia induc ion, anes hesia eme gence, su gical lead-
in, o su gical lead-ou las ing longe han 180 minu es. We
de ined his plausibili y check oge he wi h digmed GmbH.
Mo e de ailed plausibili y checks would equi e signi ican ly
g ea e e o and medical expe ise, especially o p ocess
imes ha a e oo sho . No e ha we allow o all s ic ly
posi i e incision- o-closu e du a ions. Howe e , we exclude
all su ge ies wi h a closu e- o-incision ime less han o equal
o ze o, using his seconda y p ocess ime o an addi ional
plausibili y check. We also emo ed all su ge ies wi h in alid
OPS codes and all hospi als wi h less han 100 da a poin s
in he emaining da a se . As men ioned in Sec ion 3.2,we
exclude all su ge ies wi h local anes hesia explici ly eco ded
and assume he anes hesia o all emaining su ge ies o be no
local.
We use one-way ANOVA o de e mine whe he we should
use u gency (“elec i e” s. “eme gency”) o ype o su gical
pa ien (“inpa ien ” s. “ou pa ien ”) as addi ional g ouping
pa ame e s o case mix de ini ion besides hospi al LOC and
su gical special y. Fo his, we andomly choose i edi e en
combina ions o hospi al LOC, su gical special y, and OPS
code (aiming o a su icien ly high amoun o da a poin s o
each pa ame e combina ion) o each o he wo pa ame e s
and un ANOVA o he anes hesia induc ion du a ion and
he incision- o-closu e ime o each o he combina ions. 8
ou o 10 ANOVAs o he ype o su gical pa ien show a
signi ican impac o he g ouping pa ame e on he consid-
e ed p ocess du a ion (alpha=0.05). Fo u gency, 4 ou o 10
ANOVAs show signi ican esul s. Based on his, we decided
ouse he ype o su gicalpa ien asa u he g ouping pa am-
e e . Consequen ly, we emo ed all su ge ies wi h no ype o
su gical pa ien eco ded om ou da a se . We do no u he
di e en ia e based on he su ge y u gency.
To ensu e a su icien s a is ical powe o he dis ibu ions
ha we la e i indi idually o e e y unique combina ion
o hospi al LOC, su gical special y, ype o su gical pa ien ,
OPS code, and p ocess ime, we e en ually emo ed all com-
bina ions wi h class sizes o less han 30 da a poin s.
Ou inal da a selec ion which we name he main da a
se includes su gical p ocess da a om 411 su gical depa -
men s o 139 hospi als wi h a o al o 207,635 da a poin s,
wi h 1494 obse a ions pe hospi al o 505 obse a ions pe
depa men on a e age.Based onou main da a se , we de i e
wo sepa a e da a se s o de e mine he du a ion dis ibu ion
o OR cleaning (calcula ed as P10 - A9) and closu e- o-
incision ime, espec i ely. In each o he wo addi ional da a
se s, we emo e all da a poin s wi h alues o less han o
equal o ze o and g ea e han 120 min o he co espond-
ing p ocess ime, ollowing Schus e e al. [79]. We g oup
he da a using hospi al LOC and su gical special y o he
closu e- o-incision ime.7Fo OR cleaning, we addi ionally
di e en ia e based on he ype o su gical pa ien a e ca -
ying ou a co esponding ANOVA, simila o he p ocedu e
desc ibed abo e. We again emo e classes wi h less han 30
obse a ions in each da a se . Due o ou da a selec ion p o-
cedu e, no all pa ame e combina ions o he h ee g ouping
pa ame e s a e ep esen ed in he inal OR cleaning da a se
(see Table 8). The esul ing OR cleaning da a se includes
7,442 da a poin s, and he closu e- o-incision da a se has
120,197 obse a ions.
Appendix B Dis ibu ion i ing o p ocess
du a ions
Fo each unique combina ion o selec ed pa ame e s and
o each p ocess ime, we i a ( wo-pa ame e ) logno mal,
a gamma, and a Weibull dis ibu ion using maximum-
likelihood es ima ion (MLE) [52]inR[15]. The logno mal
dis ibu ionisamong hemos popula dis ibu ions o i ing
he incision- o-closu e ime and he en i e su ge y du a ion,
i.e., case ime [84,86,98]. We choose hese h ee heo e ical
dis ibu ions since hey all a e sui able o depic he ypical
p ope ies o su gical p ocess du a ions: Con inui y, posi i e
skewness, le -side boundedness (by he ze o), and igh -side
unboundedness [60]. We de e mine he dis ibu ion pa am-
e e s and he co esponding s anda d e o s o each i ed
dis ibu ion. We calcula e he Akaike In o ma ion C i e ion
(AIC) [2] o each o he h ee co esponding MLE es ima-
ions o each pa ame e combina ion and p ocess ime. We
hen chose he heo e ical dis ibu ion wi h he lowes AIC
alue as he bes i ing [22]. Ou o he 8,425 dis ibu ions
we ha e i ed, he logno mal dis ibu ion was he bes i
in 56% o he cases, he gamma dis ibu ion in 30% o he
cases, and he Weibull dis ibu ion in he emaining 14%.
Figu e 9shows in how many cases pe cen age-wise each o
he h ee dis ibu ion ypes was he bes i o each o he
p ocess imes. We obse e ha he logno mal dis ibu ion
was he bes i in mos cases o anes hesia induc ion, anes-
hesia eme gence, incision- o-closu e, su gical lead-ou and
7Using he (main) OPS code does no make sense in his case since
closu e- o-incision ime always co esponds o wo di e en su ge ies.
123
A eady- o-use su gical p ocess da a se 345
Fig. 9 Rela i e equency o
bes i dis ibu ion ype o each
p ocess ime
Fig. 10 P obabili y densi y,
cumula i e dis ibu ion, Q-Q
and P-P plo s o anes hesia
induc ion o OPS code
5-793.k6 (T auma su ge y
inpa ien s in a hospi al o e ing
specialized ca e)
123
346 G. Ko zhene ich e al.
Fig. 11 P obabili y densi y,
cumula i e dis ibu ion, Q-Q
and P-P plo s o su gical
lead-ou o OPS code 5-712.0
(Gynecology and obs e ics
inpa ien s in a uni e si y clinic)
OR cleaning. The gamma dis ibu ion was he mos common
bes i o su gical lead-in and closu e- o-incision du a ion.
In each case, we addi ionally pe o m goodness-o - i es s
o he selec ed bes - i dis ibu ion: Kolmogo o -Smi no
(K-S) [52] o ei he logno mal, gamma, o Weibull and addi-
ionally Shapi o-Wilk (S-W) on he loga i hmized da a i he
chosen dis ibu ion was logno mal [85]. We addi ionally ana-
lyze some andomly chosen dis ibu ions g aphically i he
p- alue o a calcula ed es s a is ic is below α=0,01.
Figu e 10a o d show he analyzed plo s o he densi y,
he cumula i e dis ibu ion, he Q-Q, and he P-P plo [52]
espec i ely, o he andomly chosen case o specialized
ca e, auma su ge y, inpa ien s, OPS code 5-793.k6, and
anes hesia induc ion. The es ima ed logno mal dis ibu ion
was ejec ed in his case due o he S-W es . The de ia ion
be ween he empi ical and he i ed dis ibu ion is obse -
able in he Q-Q plo in Fig. 10c and he P-P plo in Fig. 10d.
Mo eo e , in 10a, he a o emen ioned ypical p ope ies o
a su gical p ocess du a ion dis ibu ion wi h he empi ical
densi y unc ion can be obse ed.
Ou o 8,425 pe o med es ima ions wi h ou main da a
se , he selec ed i ed dis ibu ion was ejec ed in 24% o he
cases by he goodness-o - i es ing. S um e al. [85] name
se e al po en ial explana ions o he ejec ion o a i by
a goodness-o - i es , which hold especially in he case o
he logno mal dis ibu ion. The la e is o e -p opo ionally
o en ejec ed compa ed o he gamma and he Weibull dis-
ibu ions in ou case (38% e sus 8% and 3%, espec i ely).
One issue can be la ge sample sizes [85]: In ou case, hey
ange om 30 o 4395 obse a ions pe class, wi h 87% o
all classes ha ing a sample size equal o o less han 200. Ou
o hese, 83% ha e passed he goodness-o - i es (s). Fo
he emaining 13% o all classes wi h a sample size g ea e
han 200, 33% o he classes ha e passed he goodness-o - i
es (s). Ano he explana ion can be he so-called ies [22],
i.e., local accumula ions o pa icula disc e e alues despi e
he i ed dis ibu ion being con inuous. In ou case, his is
especially ele an o ypically sho p ocess imes such as
su gical lead-ou since he one-minu e ounding p ecision
has a subs an ial e ec , in his case, simila ly o S um e
123
A eady- o-use su gical p ocess da a se 347
Fig. 12 P obabili y densi y,
cumula i e dis ibu ion, Q-Q
and P-P plo s o
incision- o-closu e ime o OPS
code 5-465.1 (Gene al su ge y
inpa ien s in a hospi al o e ing
basic and egula ca e)
al. [85]. Figu e 11 displays his issue o he case o uni e -
si y clinics, gynecology and obs e ics, inpa ien s, OPS code
5-712.0, and su gical lead-ou .
Fu he explana ions o ejec ion by a goodness-o - i
es , as discussed by S um e al. [85] a e da a ou lie s, which
can be a possible explana ion in ou case o he incision- o-
closu e ime, o which we explici ly did no se any uppe
limi when p ocessing he o iginal da a (see Appendix A).
This can be, o example, obse ed o basic and egula ca e,
gene al su ge y, inpa ien s, and OPS code 5-465.1 in Fig. 12.
Since we do no accoun o e e y possible g ouping pa am-
e e , such as anes hesia ype o su geon, when clus e ing da a
in o classes, ou i ing samples a e implici ly he e ogeneous,
which can be ano he eason o imp ecise dis ibu ion i ing
[85].Finally, we obse einou analyseso helogno mal dis-
ibu ion ha he K-S es ac s mo e conse a i ely han he
S-W es :41% o all ejec ed logno males ima esa e ejec ed
by bo h es s. Howe e , in he emaining 59%, he ejec ion
is only made by he S-W es , while in only 2 cases ou o
1,796, he K-S es alone was esponsible o he ejec ion o
he i .
We decide o s ick o all o he i ed dis ibu ions, e en
hose ejec ed by he goodness-o - i es (s), because ollow-
ing S um e al. [85], we do no ely on he es s alone bu use
hem in combina ion wi h g aphical analysis, whe e we ind
he i s o be su icien ly good, gi en he discussed peculia i-
ies o he sample da a. We ca y ou he dis ibu ion i ing o
he OR cleaning ime and closu e- o-incision ime sepa a ely
bu simila ly as desc ibed abo e.
Acknowledgemen s We wan o hankD .med.EnnoBialas,Managing
Di ec o o digmed GmbH (Hambu g, Ge many), and he en i e digmed
GmbH o he kindly p o ided su ge y p ocess da a om he Ge man
Ope a ing Room Benchma king ini ia i e and he eminen ly aluable
suppo du ing he da a analysis. We wan o exp ess ou g a i ude o
P o . D . S e an Nickel, Head o he Chai o Disc e e Op imiza ion and
Logis ics, which is pa o he Ins i u e o Ope a ions Resea ch (IOR)
a Ka ls uhe Ins i u e o Technology (KIT). This wo k is a con inua ion
o he Mas e ’s hesis by G igo y Ko zhene ich, which was comple ed
a he Chai unde he supe ision o P o . D . Nickel. We wan o hank
P o . D . Nickel o acili a ing he coope a ion wi h digmed GmbH and
o his ex emely aluable insigh s in o he subjec . Fu he mo e, we
wan o hank he en i e Chai o Disc e e Op imiza ion and Logis ics,
he IOR, and he KIT o p o iding a highly p o essional and suppo i e
en i onmen o scien i ic esea ch.
123
348 G. Ko zhene ich e al.
Funding The au ho s did no ecei e suppo om any o ganiza ion o
he submi ed wo k.
Da a A ailabili y Su gicalcasemixesand dis ibu ionso pe i-ope a i e
su gical p ocess du a ions o Ge man hospi als [da a se ]. 2022. Zen-
odo h ps://doi.o g/10.5281/zenodo.7147921
Decla a ions
E hical s anda d E hics app o al is no needed o his esea ch.
Con lic o in e es The au ho s ha e no ele an inancial o non-
inancial in e es s o disclose.
Open Access This a icle is licensed unde a C ea i e Commons
A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap-
a ion, dis ibu ion and ep oduc ion in any medium o o ma , as
long as you gi e app op ia e c edi o he o iginal au ho (s) and he
sou ce, p o ide a link o he C ea i e Commons licence, and indi-
ca e i changes we e made. The images o o he hi d pa y ma e ial
in his a icle a e included in he a icle’s C ea i e Commons licence,
unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial
is no included in he a icle’s C ea i e Commons licence and you
in ended use is no pe mi ed by s a u o y egula ion o exceeds he
pe mi eduse,you will need oob ainpe missiondi ec ly om hecopy-
igh holde . To iew a copy o his licence, isi h p://c ea i ecomm
ons.o g/licenses/by/4.0/.
Re e ences
1. Agency o Clinical Inno a ion (2014) Ope a ing Thea e E i-
ciency Guidelines: A guide o he e icien managemen o
ope a ing hea es in New Sou h Wales hospi als
2. Akaike H (1973) In o ma ion heo y and an ex ension o he maxi-
mum likelihood p inciple. In: P oc. 2nd In e na ional Symposium
on In o ma ion Theo y, 1973, Akademiai Kiado, pp 267–281
3. Aksamen o I, Twinanda AP, Mu e D e al (2017) Deep neu al
ne wo ks p edic emaining su ge y du a ion om cholecys ec-
omy ideos. In: In e na ional Con e ence on Medical Image
Compu ing and Compu e -Assis ed In e en ion, Sp inge , pp
586–593. h ps://doi.o g/10.1007/978-3-319-66185-8_66
4. Baesle F, Ga ica J, Co ea R (2015) Simula ion op imisa-
ion o ope a ing oom scheduling. In e na ional Jou nal o
Simula ion Modelling 14(2):215–226. h ps://doi.o g/10.2507/
IJSIMM14(2)3.287
5. Ba un S, Den on BT, Huschka TR e al (2011) Ope a ing
oom pooling and pa allel su ge y p ocessing unde unce ain y.
INFORMS J Compu 23(2):220–237. h ps://doi.o g/10.1287/
ijoc.1100.0396
6. Baue M, Hinz J, Klockge he -Radke A (2010) Gö inge Lei -
aden ü OP-Manage (eng. "Gö ingen guidelines o OR man-
age s"). De Anaes hesis 59(1):69–79. h ps://doi.o g/10.1007/
s00101-009-1662-5
7. Baue M, Auhube T, K aus R e al (2020) The ge man pe iope a-
i e p ocedu al ime glossa y (2020 edi ion) a join ecommenda-
ion by he bda, bdc, opm, opmo, oga i and s opm. ANASTHE-
SIOLOGIE&INTENSIVMEDIZIN61:516–531.h ps://doi.o g/
10.19224/ai2020.516
8. den Be gh JV, Beliën J, B uecke PD e al (2013) Pe sonnel
scheduling: A li e a u e e iew. Eu J Ope Res 226:367–385.
h ps://doi.o g/10.1016/j.ejo .2012.11.029
9. Bialas E, Schus e M, Taube C e al (2014) Fün Jah e OP-
P ozessda en Benchma king (2009–2013): De ak uelle S and
des P og amms on VOPM, DGAI/BDA und BDC (eng. “Fi e
yea s o benchma king OR p ocess da a (2009–2013): The cu -
en s a us o he VOPM, DGAI/BDA and BDC p og amme").
Anäs hesiologie & In ensi medizin 55:594–613
10. Blake JT, Ca e MW (2002) A goal p og amming app oach o
s a egic esou cealloca ionin acu eca ehospi als.Eu JOpe Res
140:541–561. h ps://doi.o g/10.1016/S0377-2217(01)00219-3
11. Boggs SD, Tsai MH, U man RD (2018) The associa ion o
anes hesia clinical di ec o s (aacd) glossa y o imes used o
scheduling and moni o ing o diagnos ic and he apeu ic p o-
cedu es. J Med Sys 42(9):171. h ps://doi.o g/10.1007/s10916-
018-1022-6
12. B own MJ, Sub amanian A, Cu y TB e al (2014) Imp o ing
ope a ing oom p oduc i i y ia pa allel anes hesia p ocessing.
In J Heal h Ca e Qual Assu . h ps://doi.o g/10.1108/IJHCQA-
11-2013-0129
13. Bundesam ü S a is ik, Schweiz. Gesundhei swesen, Spi äle .
(Las accessed on 28.12.2022) h ps://www.b s.admin.ch/b s/de/
home/s a is iken/gesundhei /gesundhei swesen/spi aele .h ml
14. Ca doen B, Demeulemees e E, Beliën J (2010) Ope a ing oom
planning and scheduling: A li e a u e e iew. Eu J Ope Res
201(3):921–932. h ps://doi.o g/10.1016/j.ejo .2009.04.011
15. Deligne e-Mulle ML, Du ang C e al (2015) i dis plus: An
package o i ing dis ibu ions. Jou nal o s a is ical so wa e
64(4):1–34. h ps://doi.o g/10.18637/jss. 064.i04
16. Demeulemees e E, Beliën J, Ca doen B e al (2013) Ope a ing
Room Planning and Scheduling, Sp inge New Yo k, New Yo k,
NY, pp 121–152. h ps://doi.o g/10.1007/978-1-4614-5885-2_5
17. Demi kol E, Meh a S, Uzsoy R (1998) Benchma ks o shop
scheduling p oblems. Eu J Ope Res 109(1):137–141. h ps://
doi.o g/10.1016/S0377-2217(97)00019-2
18. Dex e F, T aub RD (2002) How o schedule elec i e su gical
cases in o speci ic ope a ing ooms o maximize he e iciency o
use o ope a ing oom ime. Anes h Analg 94:933–942. h ps://
doi.o g/10.1097/00000539-200204000-00030
19. Dex e F, Dex e EU, Ledol e J (2010) In luence o p ocedu e
classi ica ion on p ocess a iabili y and pa ame e unce ain y o
su gical case du a ions. Anes h Analg 110(4):1155–1163. h ps://
doi.o g/10.1213/ANE.0b013e3181d3e79d
20. digmedGmbH:Homepage.(Las accessedon28.12.2022)h ps://
www.digmed.de/
21. digmed GmbH: Publika ionen. (Las accessed on 28.12.2022)
h ps://www.digmed.de/publika ionen/
22. Do mann CF (2017) Pa ame ische S a is ik, 2., übe a bei e e und
e wei e e Au lage (eng. "Pa ame ical S a is ics, 2., modi ied and
ex ended edi ion"). Sp inge Spek um Be lin, Heidelbe g
23. E ns A, Jiang H, K ishnamoo hy M e al (2004) S a schedul-
ing and os e ing: A e iew o applica ions, me hods and models.
Eu opean Jou nal o Ope a ional Resea ch 153:3–27. h ps://doi.
o g/10.1016/S0377-2217(03)00095-X, ime abling and Ros e -
ing
24. Fede al Ins i u e o D ugs and Medical De ices (B A M):
OPS. (Las accessed on 28.12.2022) h ps://www.b a m.
de/EN/Code-sys ems/Classi ica ions/OPS-ICHI/OPS/_node.
h ml;jsessionid=06F302DC5F77555C812176F179338333.
in ane 251
25. Fos e T (2012) Da a o benchma king you o ’s pe o mance.
OR Manage 28(1):1–5
26. F ühwi h-Schna e S (2006) Fini e Mix u e and Ma ko Swi ch-
ing Models. Sp inge , New Yo k,. h ps://doi.o g/10.1007/978-0-
387-35768-3
27. gesund.bund.de: OPS code 5-870. (Las accessed on 28.12.2022)
h ps://gesund.bund.de/en/ops-code-sea ch/5-870
123
A eady- o-use su gical p ocess da a se 349
28. gesund.bund.de: OPS code 5-870.61. (Las accessed on
28.12.2022) h ps://gesund.bund.de/en/ops-code-sea ch/5-870-
61
29. Gesundhei .g .a . Im K ankenhaus: die s a ionä e Ve -
so gung. (eng. "Inside a hospi al: The inpa ien ca e.")
(Las accessed on 28.12.2022) h ps://www.gesundhei .
g .a /gesundhei sleis ungen/k ankenhausau en hal /sys em-
k ankenhaus
30. Gomes C, Almada-Lobo B, Bo ges J e al (2012) In eg a ing da a
mining and op imiza ion echniques on su ge y scheduling. In e -
na ional Con e ence on Ad anced Da a Mining and Applica ions
pp 589–602. h ps://doi.o g/10.1007/978-3-642-35527-1_49
31. Gue ie o F, Guido R (2011) Ope a ional esea ch in he manage-
men o he ope a ing hea e: a su ey. Heal h Ca e Manag Sci
14(1):89–114. h ps://doi.o g/10.1007/s10729-010-9143-6
32. Guido R, Con o i D (2017) A hyb id gene ic app oach o sol -
ing an in eg a ed mul i-objec i e ope a ing oom planning and
scheduling p oblem. Compu e s & Ope a ions Resea ch 87:270–
282. h ps://doi.o g/10.1016/j.co .2016.11.009
33. Gul S, Den on BT, Fowle JW e al (2011) Bi-c i e ia schedul-
ing o su gical se ices o an ou pa ien p ocedu e cen e . P od
Ope Manag 20:406–417. h ps://doi.o g/10.1111/j.1937-5956.
2011.01232.x
34. Gul S, Den on BT, Fowle JW (2015) A p og essi e hedging
app oach o su ge y planning unde unce ain y. INFORMS J
Compu 27:755–772. h ps://doi.o g/10.1287/ijoc.2015.0658
35. Gü ¸S (2018) E en T (2018) Applica ion o ope a ional esea ch
echniques in ope a ing oom scheduling p oblems: Li e a u e
o e iew. Jou nal o heal hca e enginee ing 5341:394. h ps://doi.
o g/10.1155/2018/5341394
36. Hans E, Wullink G, an Houdenho en M e al (2008) Robus
su ge y loading. Eu J Ope Res 185(3):1038–1050. h ps://doi.
o g/10.1016/j.ejo .2006.08.022
37. Hans EW, Vanbe kel PT (2012) Ope a ing hea e planning and
scheduling. In: Hall R (ed) Handbook o Heal hca e Sys em
Scheduling, In e na ional Se ies in Ope a ions Resea ch & Man-
agemen Science, ol 168. Sp inge US, Bos on, MA, p 105–130,
h ps://doi.o g/10.1007/978-1-4614-1734-7_5
38. Ha is S, Claudio D (2022) Cu en ends in ope a ing oom
scheduling 2015 o 2020: a li e a u e e iew. In: Ope a ions
Resea ch Fo um, Sp inge , pp 1–42, h ps://doi.o g/10.1007/
s43069-022-00134-y
39. Ho S, Fügene A, Schoen elde J e al (2017) Case mix planning
in hospi als: a e iew and u u e agenda. Heal h Ca e Manag Sci
20:207–220. h ps://doi.o g/10.1007/s10729-015-9342-2
40. Holmg en J, Pe sson M (2016) An op imiza ion model o
sequence dependen pa allel ope a ing oom scheduling. In:
Heal h ca e sys ems enginee ing o scien is s and p ac i ione s.
Sp inge , p 41–51, h ps://doi.o g/10.1007/978-3-319-35132-
2_5
41. Hulsho PJH, Ko beek N, Bouche ie RJ e al (2012) Taxonomic
classi ica ion o planning decisions in heal h ca e: a s uc u ed
e iew o he s a e o he a in o /ms. Heal h Sys ems 1:129–175.
h ps://doi.o g/10.1057/hs.2012.18
42. Jous a P, Mees e R, an Ophem H (2013) Can s a is icians bea
su geons a he planning o ope a ions? Empi ical Economics
44(3):1697–1718. h ps://doi.o g/10.1007/s00181-012-0594-0
43. JR OP-TIMIERT GmbH: Wa um wi Ihnen je z einen neuen
OP-Benchma k anbie en. (eng. "JR OP-TIMIERT GmbH: Why
we o e you a new OR benchma k now.") (Las accessed
on 28.12.2022) h ps://www.j -op imie .de/wa um-wi -ihnen-
je z -einen-neuen-op-benchma k-anbie en/
44. Jung KS, Pinedo M, S iskanda ajah C e al (2019) Scheduling
elec i e su ge ies wi h eme gency pa ien s a sha ed ope a ing
ooms. P od Ope Manag 28(6):1407–1430. h ps://doi.o g/10.
1111/poms.12993
45. Ka aca O, Baue M, Taube C e al (2019) Ko elie die Leis ungs-
menge eines K ankenhauses mi de chi u gischen P ozesszei ?
: Re ospek i e Analyse de jeweils ün häu igs en P oze-
du en aus den Be eichen Visze alchi u gie, Un all-/o hopädische
Chi u gie und Gynäkologie/Gebu shil e aus dem Benchma king-
P og amm on BDA, BDC und VOPM (eng. "Does hospi al
olume co ela e wi h su gical p ocess ime? Re ospec i e anal-
ysis o he i e mos common p ocedu es o isce al su ge y,
auma and o hopedic su ge y and gynecology/obs e ics om
he benchma king p og am o he Be u s e band Deu sche Anäs-
hesis en (BDA), Be u s e band Deu sche Chi u gen (BDC)
and Ve band ü OP-Managemen (VOPM)"). De Anaes hesis
68(4):218–227. h ps://doi.o g/10.1007/s00101-019-0559-1
46. KHZV Rheinland: OP-Benchma king. (Las accessed on
28.12.2022) h ps://web.a chi e.o g/web/20220524002946/,
h p://www.khz -op.de/index.php
47. Kolisch R, Sp eche A, D exl A (1995) Cha ac e iza ion and
gene a ion o a gene al class o esou ce-cons ained p ojec
scheduling p oblems. Manage Sci 41(10):1693–1703. h ps://doi.
o g/10.1287/mnsc.41.10.1693
48. Ko zhene ich G, Zande A (2022) Su gical case mixes and dis-
ibu ions o pe iope a i e su gical p ocess du a ions o ge man
hospi als [da a se ]. h ps://doi.o g/10.5281/zenodo.7147921
49. Kougias P, Tiwa i V, Be ge DH (2016) Use o simula ion o
assess a s a is ically d i en su gical scheduling sys em. J Su g
Res 201(2):306–312. h ps://doi.o g/10.1016/j.jss.2015.10.043
50. Landa P, A inghie i R, So iano P e al (2016) A hyb id op imiza-
ion algo i hm o su ge ies scheduling. Ope a ions Resea ch o
Heal h Ca e 8:103–114. h ps://doi.o g/10.1016/j.o hc.2016.01.
001
51. La o e-Núñez G, Lüe -Villag a A, Ma iano V e al (2016)
Scheduling ope a ing ooms wi h conside a ion o all esou ces,
pos anes hesia beds and eme gency su ge ies. Compu e s &
Indus ial Enginee ing 97:248–257. h ps://doi.o g/10.1016/j.cie.
2016.05.016
52. Law AM (2015) Simula ion modeling and analysis, i h, edi ion.
McG aw-Hill se ies in indus ial enginee ing and managemen
science, McG aw-Hill Educa ion, Dubuque
53. Lee ink G, Hans EW (2018) Case mix classi ica ion and a bench-
ma k se o su ge y scheduling. J Sched 21(1):17–33. h ps://doi.
o g/10.1007/s10951-017-0539-8
54. Leh onen JM, To kki P, Pel oko pi A e al (2013) Inc easing ope -
a ing oom p oduc i i y by du a ion ca ego ies and a news endo
model. In J Heal h Ca e Qual Assu 26(2):80–92. h ps://doi.o g/
10.1108/09526861311297307
55. Lenge sdo B, Liehn M, S einmülle L e al (2021) G undlagen
(eng. "Basics"). In: Liehn M, Lenge sdo B, S einmülle L, e al.
(eds) OP-Handbuch. Sp inge Be lin Heidelbe g, Be lin, Heidel-
be g, p 1–35, h ps://doi.o g/10.1007/978-3-662-61101-2_1
56. Maca io A (2009) T u h in scheduling: is i possible o
accu a ely p edic how long a su gical case will las ?
Anes h Analg 108(3):681–685. h ps://doi.o g/10.1213/ane.
0b013e318196a617
57. Makboul S, Kha aja S, Abbassi A e al (2022) A wo-s age obus
op imiza ion app oach o he mas e su gical schedule p ob-
lem unde unce ain y conside ing downs eam esou ces. Heal h
Ca e Manag Sci 25(1):63–88. h ps://doi.o g/10.1007/s10729-
021-09572-2
58. Ma ques I, Cap i o ME, Vaz Pa o M (2015) A bic i e ia heu is ic
o an elec i e su ge y scheduling p oblem. Heal h Ca e Manag
Sci 18(3):251–266. h ps://doi.o g/10.1007/s10729-014-9305-z
59. Meie J, Die z A, Boehm A e al (2015) P edic ing ea men
p ocess s eps om e en s. J Biomed In o m 53:308–319. h ps://
doi.o g/10.1016/j.jbi.2014.12.003
60. Messe C (2014) Simula ionsmodell zu Planung de Anzahl an
Schleusen in K ankenhäuse n (eng. "Simula ion model o plan-
123
350 G. Ko zhene ich e al.
ning he numbe o ans e ooms in hospi als"). Mas e hesis,
Ka ls uhe Ins i u e o Technology, Ka ls uhe, Ge many
61. Messe C, Zande A, A nolds IV e al (2015) Wie iele Pa ien-
enschleusen b auch mein OP-Be eich? : Ein luss de Zahl on
OP-Schleusen au Wa ezei en und Pa ien endu chsa z im OP -
Analyse mi hil e eine Simula ion (eng. "How many pa ien ans-
e ooms a e necessa y o my OR sui e? E ec o he numbe o
OR ans e ooms on wai ing imes and pa ien h oughpu in he
OR - analysis by simula ion"). De Anaes hesis 64(12):958–967.
h ps://doi.o g/10.1007/s00101-015-0108-5
62. Molina-Pa ien e JM, Fe nandez-Viagas V, F aminan JM (2015)
In eg a ed ope a ing oom planning and scheduling p oblem
wi h assis an su geon dependen su ge y du a ions. Compu e s
& Indus ial Enginee ing 82:8–20. h ps://doi.o g/10.1016/j.cie.
2015.01.006
63. Neumann J, Ang ick C, Höhn C e al (2020) Su gical wo k low
simula ion o he design and assessmen o ope a ing oom se ups
in o hopedic su ge y. BMC Med In o m Decis Mak 20(1):1–20.
h ps://doi.o g/10.1186/s12911-020-1086-3
64. Neumu h T, Jannin P, Schlombe g J e al (2011) Analysis o
su gical in e en ion popula ions using gene ic su gical p ocess
models. In J Compu Assis Radiol Su g 6(1):59–71. h ps://doi.
o g/10.1007/s11548-010-0475-y
65. NHS Benchma king Ne wo k: 2018 Ope a ing Thea es P ojec
- Resul s published. (Las accessed on 28.12.2022) h ps://www.
nhsbenchma king.nhs.uk/news/2018-ope a ing- hea es- esul s-
published
66. NHS Benchma king Ne wo k: Acu e Sec o . (Las accessed on
28.12.2022) h ps://www.nhsbenchma king.nhs.uk/acu e-sec o
67. Ozen A, Ma mo Y, Rohlede T e al (2016) Op imiza ion and
simula ion o o hopedic spine su ge y cases a mayo clinic. Man-
u ac u ing & Se ice Ope a ions Managemen 18(1):157–175.
h ps://doi.o g/10.1287/msom.2015.0564
68. Ped on S, Win e V, Oppel EM e al (2017) Ope a ing oom e i-
ciency be o e and a e en ance in a benchma king p og am o
su gical p ocess da a. J Med Sys 41(10):151. h ps://doi.o g/10.
1007/s10916-017-0798-0
69. Pham DN, Klinke A (2008) Su gical case scheduling as a gene -
alized job shop scheduling p oblem. Eu J Ope Res 185(3):1011–
1025. h ps://doi.o g/10.1016/j.ejo .2006.03.059
70. Rahimi I, Gandomi AH (2021) A comp ehensi e e iew and
analysis o ope a ing oom and su ge y scheduling. A chi es o
Compu a ional Me hods in Enginee ing. h ps://doi.o g/10.1007/
S11831-020-09432-2
71. Reimbu semen Ins i u e, Glossa : Fachab eilungsschlüssel. (eng.
"Special y index") (Las accessed on 28.12.2022) h ps://
eimbu semen .ins i u e/glossa / achab eilungsschluessel/
72. Reimbu semen Ins i u e: OPS Basics. (Las accessed on
28.12.2022) h ps:// eimbu semen .ins i u e/en/ops-applica ion/
73. Rieke M, P emm M, Klein A e al (2017) P edic ing he
du a ion o su ge ies o imp o e p ocess e iciency in hospi als.
In: Twen y-Fi h Eu opean Con e ence on In o ma ion Sys ems
(ECIS), Guima ães, Po ugal, 2017
74. RiiseA,ManninoC,Bu keEK(2016)Modellingandsol ing gen-
e alised ope a ional su ge y scheduling p oblems. Compu e s &
Ope a ionsResea ch66:1–11.h ps://doi.o g/10.1016/j.co .2015.
07.003
75. Samud a M, Rie CV, Demeulemees e E e al (2016) Scheduling
ope a ing ooms: achie emen s, challenges and pi alls. J Sched.
h ps://doi.o g/10.1007/S10951-016-0489-6
76. Schul z J, Claudio D (2014) Va iabili y based su gical scheduling:
A simula ion app oach. In: P oceedings o he Win e Simula ion
Con e ence 2014. IEEE, pp 1353–1364, h ps://doi.o g/10.1109/
WSC.2014.7019990
77. Schus e M, Wicha LL, Fiege M (2007) Kennzahlen de OP-
E izienz. My hos und E idenz de S eue ungskennzahlen im
OP-Managemen (eng. "Key pe o mance indica o s o OR e i-
ciency. My hs and e idence o key pe o mance indica o s in OR
managemen "). De Anaes hesis 56(3):259–271. h ps://doi.o g/
10.1007/s00101-006-1126-0
78. Schus e M, Pezzella M, Taube C e al (2013) Delays in s a -
ing mo ning ope a ing lis s: an analysis o mo e han 20,000
cases in 22 ge man hospi als. Deu sches A z ebla in e na ional
110(14):237–243. h ps://doi.o g/10.3238/a z ebl.2013.0237
79. Schus e M, Be heau S, Taube C e al (2014) Übe lappende
Anäs hesie-Einlei ungen und pe iope a i e Wechselzei en (eng.
“O e lapping anaes hesia induc ion and pe iope a i e u no e
imes - An analysis o he benchma k p og am me o he
BDA/BDC and VOPM"). Anäs hesiologie & In ensi medizin
55:654–661
80. Shylo OV, P okopye OA, Schae e AJ (2013) S ochas ic ope -
a ing oom scheduling o high- olume special ies unde block
booking. INFORMS J Compu 25(4):682–692. h ps://doi.o g/10.
1287/ijoc.1120.0530
81. SINTEF: Op imiza ion in Heal h Ca e. (Las accessed on
28.12.2022) h ps://www.sin e .no/en/digi al/depa men s-new/
applied-ma hema ics/op imiza ion/heal h-ca e-op imiza ion/
82. S a e o Queensland (2017) Ope a ing Thea e E iciency - Guide-
line. h ps://www.heal h.qld.go .au/__da a/asse s/pd _ ile/0022/
640138/qh-gdl-443.pd
83. S a is isches Bundesam . Gesundhei , K ankenhäuse . (Las
accessed on 28.12.2022) h ps://www.des a is.de/DE/Themen/
Gesellscha -Umwel /Gesundhei /K ankenhaeuse /_inhal .h ml
84. S epaniak PS, Heij C, Mannae s GHH e al (2009) Modeling
p ocedu e and su gical imes o cu en p ocedu al e minology-
anes hesia-su geon combina ions and e alua ion in e ms o case-
du a ion p edic ion and ope a ing oom e iciency: a mul icen e
s udy. Anes h Analg 109(4):1232–1245. h ps://doi.o g/10.1213/
ANE.0b013e3181b5de07
85. S um DP, May JH, Va gas LG (2000) Modeling he unce -
ain y o su gical p ocedu e imes: compa ison o log-no mal and
no mal models. The Jou nal o he Ame ican Socie y o Anes he-
siologis s 92(4):1160–1167. h ps://doi.o g/10.1097/00000542-
200004000-00035
86. S um DP, Sampson AR, May JH e al (2000) Su geon and ype
o anes hesia p edic a iabili y in su gical p ocedu e imes. The
Jou nalo heAme icanSocie y o Anes hesiologis s 92(5):1454–
1466. h ps://doi.o g/10.1097/00000542-200005000-00036
87. Ull ich M, Weise T, Awas hi A e al (2018) A gene ic p ob-
lem ins ance gene a o o disc e e op imiza ion p oblems. In:
P oceedings o he Gene ic and E olu iona y Compu a ion
Con e ence Companion, pp 1761–1768, h ps://doi.o g/10.1145/
3205651.3208284
88. Uni e si y o Twen e, CHOIR: Su ge y Scheduling Benchma k
Se . (Las accessed on 28.12.2022) h ps://www.u wen e.nl/en/
choi / esea ch/Benchma kORScheduling/
89. an Essen JT, Hans EW, Hu ink JL e al (2012) Minimizing he
wai ing ime o eme gency su ge y. Ope a ions Resea ch o
Heal h Ca e 1(2–3):34–44. h ps://doi.o g/10.1016/j.o hc.2012.
05.002
90. an Veen-Be kx E, Bi e J, Elkhuizen SG e al (2014) The in lu-
ence o anes hesia-con olled ime on ope a ing oom scheduling
in Du ch uni e si y medical cen es. Canadian jou nal o anaes-
hesia= Jou nalcanadien d’anes hesie 61(6):524–532. h ps://doi.
o g/10.1007/s12630-014-0134-9
91. an Veen-Be kx E, de Ko ne DF, Oli ie OS e al (2016) Bench-
ma king ope a ing oom depa men s in he Ne he lands. Bench-
ma king: An In e na ional Jou nal 23(5):1171–1192. h ps://doi.
o g/10.1108/BIJ-04-2014-0035
92. Vanbe kel PT, Bouche ie RJ, Hans EW e al (2011) An exac
app oach o ela ing eco e ing su gical pa ien wo kload o he
123
A eady- o-use su gical p ocess da a se 351
mas e su gical schedule. Jou nal o he Ope a ional Resea ch
Socie y 62:1851–1860. h ps://doi.o g/10.1057/jo s.2010.141
93. Vanhoucke M, Maenhou B (2007) Nsplib–a nu se scheduling
p oblem lib a y: A ool o e alua e (me a-) heu is ic p ocedu es.
In: Ope a ional esea ch o heal h policy: making be e deci-
sions, p oceedings o he 31s annual mee ing o he wo king
g oup on ope a ions esea ch applied o heal h se ices, pp 151–
165
94. Wang L, Demeulemees e E, Vans eenkis e N e al (2021) Ope a -
ing oom planning and scheduling o ou pa ien s and inpa ien s:
A e iew and u u e esea ch. Ope a ions Resea ch o Heal h
Ca e 31(100):323. h ps://doi.o g/10.1016/j.o hc.2021.100323
95. Wikipedia:Ve so gungss u e (eng. "Le elo ca e")(Las accessed
on 28.12.2022) h ps://de.wikipedia.o g/wiki/Ve so gungss u e
96. Wullink G, Van Houdenho en M, Hans EW e al (2007) Clos-
ing eme gency ope a ing ooms imp o es e iciency. J Med Sys
31(6):543–546. h ps://doi.o g/10.1007/s10916-007-9096-6
97. Zhang J, D idi M, El Moudni A (2017) A s ochas ic sho es -pa h
mdp model wi h dead ends o ope a ing ooms
planning. In: 2017 23 d In e na ional Con e ence on Au oma ion
and Compu ing (ICAC), IEEE, pp 1–6, h ps://doi.o g/10.23919/
IConAC.2017.8081974
98. Zhou J, Dex e F (1998) Me hod o assis in he scheduling o
add-on su gical cases-uppe p edic ion bounds o su gical case
du a ions based on he log-no mal dis ibu ion. The Jou nal o he
Ame ican Socie y o Anes hesiologis s 89(5):1228–1232. h ps://
doi.o g/10.1097/00000542-199811000-00024
99. Zhu S, Fan W, Yang S e al (2019) Ope a ing oom planning and
su gical case scheduling: a e iew o li e a u e. J Comb Op im
37(3):757–805. h ps://doi.o g/10.1007/s10878-018-0322-6
100. Zonde land ME, B inkho S, No ing IC e al (2021) A ma ko
modelling app oach o su gical p ocess analysis in ca a ac
su ge y. In: Handbook o Heal hca e Logis ics. Sp inge , p 97–
110, h ps://doi.o g/10.1007/978-3-030-60212-3_6
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju is-
dic ional claims in published maps and ins i u ional a ilia ions.
123
