• Nenhum resultado encontrado

Prospective assessment of the risk of postoperative pulmonary complications in patients submitted to upper abdominal surgery

N/A
N/A
Protected

Academic year: 2017

Share "Prospective assessment of the risk of postoperative pulmonary complications in patients submitted to upper abdominal surgery"

Copied!
10
0
0

Texto

(1)

Eanes Delgado Barros Pereira, Ana Luisa Godoy Fernandes, Meide da Silva Anção, Clóvis de Araújo Peres, Álvaro Nagib Atallah, Sonia Maria Faresin

Prospe ctive asse ssme nt of the risk of

postope rative pulmonary complications in

patie nts submitte d to uppe r abdominal surge ry

Pulmonary Division, Universidade Federal de São Paulo, São Paulo, Brazil

INTRODUCTION

The re la tio nship b e tw e e n p re o p e ra tive variables and PPCs in surg ical patients has been the subject o f numero us studies. Despite recent a d va nc e s in p re o p e ra tive ma na g e me nt, po sto pe ra tive re spira to ry mo rb idity is still a c o mmo n pro b lem, espec ia lly fo llo wing upper abdo minal surg ery.1 -6

The ma in risk fa c to rs tha t ha ve b e e n a sso c ia te d w ith PPC s a re : smo king , c hro nic o bstructive pulmo nary disease, advanced ag e, site a nd d ura tio n o f surg e ry, o b e sity a nd c o mo rb id ity.7 -1 3 The p re o p e ra tive e va lua tio n sho uld inc lude ste ps to pre pa re pa tie nts fo r surg ery a nd to identify tho se a t hig h risk fo r d e ve lo p ing c o mp lic a tio ns, thus a llo w ing physic ia ns to ta ke pro phyla c tic me a sure s to reduce the incidence o f PPCs.

The incidence o f PPCs varies fro m 1 0 to 8 0 %.8 -1 0 This wide rang e is due to the lack o f explicitly standardiz ed definitio n o f PPCs that w o uld re d uc e inte r-o b se rve r d isa g re e me nt. Usua lly a c o mp lic a tio n is d e fine d a s a n unexpected seco nd disease entity that requires sp e c ia l tre a tme nt, w he re a s a find ing is a n abno rmality that results fro m an investig atio n. The mo st c o mmo ns PPC s a re : a te le c ta sis,

ABSTRACT

Objective: To investig ate asso ciatio ns between

preo perative variables and po sto perative pulmo nary co mplicatio ns (PPC) in elective upper abdo minal surg ery.

Design: Pro spective clinical trial.

Setting: A tertiary university ho spital.

Pa tients: 4 0 8 patients were pro spectively analyzed

during the preo perative perio d and fo llo wed up po sto peratively fo r pulmo nary co mplicatio ns.

M ea surements: Patient characteristics, with clinical and

physical evaluatio n, related diseases, smo king habits, and duratio n o f surg ery. Preo perative pulmo nary functio n tests (PFT) were perfo rmed o n 2 4 7 patients.

Results: The po sto perative pulmo nary co mplicatio n rate

was 1 4 percent. The sig nificant predicto rs in univariate analyses o f po sto perative pulmo nary co mplicatio ns were: ag e >5 0 , smo king habits, presence o f chro nic pulmo nary disease o r respirato ry sympto ms at the time o f evaluatio n, duratio n o f surg ery >2 1 0 minutes and co mo rbidity (p <0 .0 4 ). In a lo g istic reg ressio n analysis, the statistically sig nificant predicto rs were: presence o f chro nic pulmo nary disease, surg ery lasting >2 1 0 and co mo rbidity (p <0 .0 0 9 ).

Conclusions: There were three majo r clinical risk facto rs

fo r pulmo nary co mplicatio ns fo llo wing upper abdo minal surg ery: chro nic pulmo nary disease, co mo rbidity, and surg ery lasting mo re than 2 1 0 minutes. Tho se patients with three risk facto rs were three times mo re likely to develo p a PPC co mpared to patients witho ut any o f these risk facto rs (p <0 .0 0 1 ). PFT is indicated when there are uncertainties reg arding the patient’s pulmo nary status.

Key w ords: abdo minal surg ery, risk facto r, mo rbidity

Abbrevia tions: BMI = bo dy mass index; FEV1/ FVC =

fo rced expirato ry vo lume in the first seco nd divided by fo rced vital capacity; PPC = po sto perative pulmo nary co mplicatio n.

(2)

respirato ry infectio ns, bro ncho co nstrictio n and respirato ry failure.3 ,1 1 ,1 2

Preo perative spiro metric tests have been re p o rte d to b e re lia b le p re d ic to rs o f PPC s. Ho wever, the risk o f PPCs is usually estimated in he te ro g e ne o us p o p ula tio ns a nd in va rio us surg ical pro cedures,1 ,1 3 ,1 4 making it difficult to a sc erta in the rela tio nship b etween PPCs a nd previo us spiro metric abno rmalities.

The a im o f this stud y w a s to fo llo w pro spectively a g ro up o f patients underg o ing an elective upper abdo minal surgery to identify tho se fa c to rs a sso c ia te d with a n inc re a se d risk o f develo ping PPCs using a standard preo perative eva lua tio n. In a dditio n, a sub g ro up o f these p a tie nts w a s sub mitte d to sp iro me try to inve stig a te the imp o rta nc e o f this te st a s a predicto r o f PPCs in this po pulatio n.

METHODS

This stud y w a s p e rfo rme d o n 4 0 8 c o nse c utive p a tie nts und e rg o ing e le c tive abdo minal surg ery at the Federal University o f São Paulo ’s teaching ho spital. All patients were referred fo r a preo perative assessment (between Ja nua ry 1 9 9 2 a nd De c e mb e r 1 9 9 2 ), a fte r ha ving b e e n sc he d ule d fo r e le c tive up p e r abdo minal surg ery. All were o perated o n and o bserved at the same ho spital.

O f the 4 0 8 patients, 2 0 6 were men and 2 0 2 were wo men and their mean age was 5 5 + 1 5 years. O perative pro cedures are given in Table 1 .

Preoperative Assessment

All the patients were submitted to an initial clinical evaluatio n with a standard questio nnaire fo r c linic a l histo ry a nd a c o mple te physic a l e xa mina tio n. This pro c e dure wa s pre vio usly appro ved by the Co mmittee fo r Ethical Co ntro l o f C linic a l Re se a rc h “ In A nima N o b ili” o f UN IFESP (Federal University o f São Paulo ).

Data was o btained fro m a questio nnaire o n the presence o f respirato ry sympto ms during the week preceding surg ery, as well as clinically in regard to chro nic lung disease, smo king habits, nutritio nal status and co mo rbidity presented by the patients.

In o rder to place the patients being studied into catego ries, the fo llo wing definitio ns were used: • Patient with Respiratory Symptoms: one suffering

fro m at least o ne o f the fo llo wing sympto ms: Cough - Positive answer when asked the question: “Have you got a cough at the moment?”

Chro nic Co ug h - Po sitive answer to at least o ne o f the fo llo wing questio n: “ Do yo u co ug h habitually? ” and “ Do yo u co ug h frequently during the day, fo r at least fo ur days a week and fo r at least three co nsecutive mo nths o r mo re during the year?”

Sputum - Po sitive a nswe r to the fo llo wing questio n: “ Do yo u habitually co ug h up sputum fro m the lung s?”

• Chro nic Pulmo nary Disease: pulmo nary illness o f chro nic evo lutio n, sympto ms underg o ing treatment at present, o r no t, and with diagno sis previo usly established o r made at the time o f pre-o perative evaluatio n. W ithin this heading we re inc lude d the pre se nc e o f b ro nc hitis-emphysema, chro nic o bstructive pulmo nary disease, asthma, bro nchiectasis and interstitial lung disease.1 5 -1 7

• Current Smo ker - o ne saying that he had smo ked at least o ne cigarette a day fo r mo re than a year and who was using cig arettes at that mo ment o r who had sto pped smo king less than eight weeks ago . The co nsumptio n o f cigarettes was expressed in packet-years, that is to say the pro duct between the time o f co nsumptio n in years and the number o f packets (gro ups o f 2 0 cigarettes) smo ked per day.

Table 1 - Types of surgery w ith incision above the navel mark performed on 408 patients follow ed

up for postoperative pulmonary complications

Type o f surg ery N %

G astro entero lo g ical 3 2 5 7 9 .7

G yneco lo g ical 2 3 5 .6

Abdo minal wall 2 3 5 .6

Uro lo g y 2 0 4 .9

Vascular 9 2 .2

Retro perito nium 8 2 .0

(3)

• N utritio nal Status: evaluated by the bo dy mass index (BMI), as calculated by the fo llo wing relatio n: weig ht in Kg divided by heig ht in meters squared. The po pulatio n studied was classified as dystro phic (BMI <2 1 Kg / m2 o r

3 0 Kg / m2) o r eutro phic (BMI

2 1 Kg / m2 and <3 0 Kg / m2).1 8 ,1 9

• Co mo rbidity: including the o ccurrence o f o ne o r mo re o f the fo llo wing dise a se s: a c tive systemic arterial hypertensio n, cardio patho -lo g y and diabetes mellitus.

Pulmo nary functio n tests were perfo rmed upo n 2 4 7 patients, acco rding to ATS criteria2 0 and fulfilling tho se established by Ho usto n et al3: candidates fo r upper abdo minal surg ery o ver 6 0 years o f ag e, tho se with pulmo nary disease present a t the time o f eva lua tio n, tho se with mo rbid o besity (weig ht mo re than 1 5 0 % o ver ideal bo dy weig ht), current smo kers and bearers o f respirato ry sympto ms.

After co ncluding the evaluatio n, appro priate pro phylactic measures were suggested fo r the pre-a nd po sto pe rpre-a tive pe rio ds fo r e pre-a c h ppre-a tie nt, including: interruptio n o f the smo king habit, use o f

b ro nc ho dila to rs, c o urses o f systemic stero ids, orientation for respiratory physiotherapy exercises21 a nd the use o f lo w do ses o f hepa rin fo r the pro phylaxis o f pulmo nary thro mbo embo lism.

Postoperative Assessment

Patients were acco mpanied daily during the po sto perative perio d by the same medical team which assessed the preo perative perio d, until they were discharg ed o r died. The fo llo wing pulmo nary co mplicatio ns and acute respirato ry infectio ns were mo nito red:

• Pneumo nia: presence o f pulmo nary infiltratio n o n the chest x-ray asso ciated with at least two o f the fo llo wing signs: purulent tracheo bro nchial secretio n, elevatio n o f bo dy temperature (abo ve 3 8 . 3 º C ) a nd inc re a se o f le uko c yte s in circulatio n (o ver 2 5 % abo ve the base co unt).2 2 • Tracheo bro nchitis: increase in the quantity o r

chang ing o f the co lo r o r purulent aspect o f tracheo bro nchial secretio n with no rmal chest x-ray.2 2

• Atelectasis with clinical repercussio n: evidence o f pulmo nary atelectasis o n the chest x-ray

Table 2 - Factors related to postoperative pulmonary complications

Variable Rate o f PPC χ2 p-value

CPD 3 5 / 1 2 0 (2 9 %) 3 1 .1 6 1 0 .0 0 0 1 *

Respirato ry sympto ms 4 1 / 1 8 3 (2 2 %) 1 8 .2 4 7 0 .0 0 0 1 *

Duratio n o f surg ery >2 1 0 min. 4 5 / 2 6 8 (1 7 %) 8 .2 4 8 0 .0 3 *

Co mo rbidity 3 4 / 1 7 3 (2 0 %) 7 .2 8 2 0 .0 0 7 *

Age > 5 0 years o ld 4 6 / 2 7 3 (1 7 %) 4 .6 9 4 0 .0 3 *

•Smo king (packet-years) 4 1 / 2 3 9 (1 7 %) 4 .0 8 7 0 .0 4 *

••Current Smo king 2 6 / 1 3 6 (1 9 %) 4 .0 2 0 0 .0 4 *

BMI < 2 1 Kg/ m2 and > 3 0 Kg/ m2 2 2 / 1 8 8 (1 2 %) 1 .8 0 6 0 .1

* Statistically Sig nificant (P < 0 .0 5 ); CPD: Chro nic pulmo nary disease; BMI: Bo dy mass index; •including current smo kers and

ex-smo kers; ••including tho se that ex-smo ked until surg ery

Table 3 - M ultivariate analysis of the three main risk factors for postoperative pulmonary complications

Effect Estimate Standard erro r χ2 p-value O R (9 5 % CI)

Intercept -1 .7 9 2 1 0 .1 9 7 0 7 0 .2 9 <0 .0 0 0 1 *

CPD -0 .8 0 9 5 0 .1 5 2 7 2 8 .1 0 <0 .0 0 0 1 * 2 .2 4 (1 .6 6 -2 .9 7 )

DS >2 1 0 min. -0 .4 6 2 2 0 .1 6 6 7 7 .6 9 0 .0 0 5 * 1 .5 8 (1 .1 5 -2 .1 8 )

Co mo rbidity -0 .3 9 8 7 0 .1 5 2 6 6 .8 9 0 .0 0 9 * 1 .4 8 (1 .1 0 -1 .9 7 )

(4)

asso ciated with acute respirato ry sympto ms. • A c ute re sp ira to ry fa ilure : c linic a l p ic ture resulting fro m acutely deficient exchang e o f g a se s in the lung , ma king me c ha nic a l ventilatio n necessary fo r treatment.

• Pro lo ng e d o ro tra c he a l tub ing : ne e d fo r o ro tracheal tubing fo r mo re than 4 8 ho urs, due to maintaining mechanical ventilatio n fo r the treatment o f acute respirato ry failure o r the aspiratio n o f tracheo bro nchial secretio n in tho se unable to eliminate it spo ntaneo usly. • Pro lo ng ed mechanical ventilatio n: need fo r mechanical ventilatio n fo r mo re than 4 8 ho urs fo r the treatment o f acute respirato ry failure. • Bro ncho co nstrictio n: presence o f wheez ing

asso ciated with acute respirato ry sympto ms and use o f bro ncho dilato r medicatio n. In o rder to dismiss a diag no sis o f pulmo nary embo lism and pulmo nary edema the lung pulmo nary scan, pulmo nary arterio graphy, measurements o f c a rdia c de b it a nd pulmo na ry c a pilla ry pressure were o btained fo r all the patients who did no t present a previo us histo ry o f asthma o r c hro nic o b struc tive pulmo na ry dise a se . Bro ncho co nstrictio n related to intubing and extubing o f patients was no t co nsidered as a pulmo nary co mplicatio n.

W he n p a tie nts d ie d , the ma in a nd seco nda ry ca uses o f the fa ta l evo lutio n were determined. The data fro m the po st-mo rtem were co nsidered and in the absence o f this, clinical and labo rato ry data were used.

Statistical Methods.

Initially, a univariate analysis using a chi-square test was perfo rmed to co mpare each risk facto r (independent variable) to PPCs, as well as the odds ratio with a 95 percent confidence interval. The stepwise lo g istic reg ressio n metho d was used fo r identifying the risk facto rs. In the stepwise pro cedure, the best (mo st sig nificant F statistic) sing le risk facto r was selected fo r entry into the mo del first. Then it was determined whether the additio n o f the remaining risk facto rs increased the ability o f the mo del to predict the risk o f an o c c urre nc e o f PPC . If the a d d e d risk fa c to r impro ved the predictio n, it remained part o f the mo del; if no t, it was rejected. The alg o rithm was co mplete when no further impo rtant risk facto rs co uld be added to the mo del. N ext, the predictive mo del co efficients and the po ssible risk facto r co mbinatio ns were co mputed. These predictio ns were then co mpared with the o bserved adverse o utc o me ra te s b y using va rio us disc re pa nc y measurements.2 3 -2 5

RESULTS

O f the 4 0 8 patients studied, 2 0 6 (5 0 .4 %) were men and 2 0 2 (4 9 .5 %) were wo men. The averag e ag e o f the g ro up was 5 5 years (SD 1 5 ). The averag e ag e fo r men was 5 6 years (SD 1 4 ), rang ing fro m 1 7 to 9 0 years, while the averag e ag e fo r wo men was 5 5 years (SD 1 6 ), rang ing fro m 1 7 to 8 4 years.

The averag e duratio n o f surg ery was 2 5 2 minutes (SD 1 0 5 ). In 1 4 0 patients (3 4 %) the surg ery leng th was less than 2 1 0 minutes, while fo r 2 6 8 (6 6 %) it lasted mo re than 2 1 0 minutes. Table 1 sho ws patients classified acco rding to the type o f surg ical pro cedure carried o ut.

Amo ng the cases studied, 1 8 3 (4 5 %) were pa tients with respira to ry sympto ms, the mo st co mmo n sympto m being a co ug h, fo llo wed by

Table 4 - The distribution of 408 patients according to the presence or absence of three risk factors

(respiratory disease, comorbidity and surgical duration > 210 min) in relation to postoperative

respiratory complications

Risk facto r PPC To tal

present absent

po sitive 1 3 1 6 2 9

neg ative 1 6 2 6 3

Total 14 78 92

sensitivity = 9 3 % (9 5 % CI 7 3 % to 9 1 %)

specificity = 7 9 % (9 5 % CI 8 0 % to 1 0 0 %)

po sitive predictive value = 4 5 % (9 5 % CI 2 9 % to 6 7 %)

negative predictive value = 9 8 % (9 5 % CI 9 5 % to 1 0 0 %)

prevalence o f PPC = 1 5 % (9 5 % CI 8 % to 2 4 %)

po sitive likeliho o d ratio 5 .0 5 (9 5 % CI 0 .7 6 to 3 3 .4 9 )

(5)

expecto ratio n and wheezing .

Chro nic pulmo nary disease was diag no sed in 1 2 0 patients (2 9 %), the mo st co mmo n being c hro nic o b struc tive p ulmo na ry d ise a se a nd asthma. There were 1 3 6 current smo kers (3 3 %) with an averag e o f 2 7 packet-years (SD 1 7 ). The averag e bo dy mass index was 2 4 .8 2 (SD 5 .9 0 ) and in acco rdance with this, 2 2 0 eutro phic p a tie nts (5 4 % ), 7 8 o b e se (1 9 % ) a nd 1 1 0 underno urished (2 7 %) were o bserved.

C o mo rb id ity a sso c ia te d w ith the b a sic surgical illness was observed in 173 patients (42%). O nly three patients (2 %) presented three associated clinical diagno ses (systemic arterial hypertensio n, diabetes mellitus and cardio patho lo gy), while 4 2 patients (2 4 %) had two clinical illness and the mo st co mmo n asso ciatio n in this case was systemic arterial hypertensio n with cardio patho lo gy (1 1 %), fo llo wed by systemic arterial hypertensio n with diabetes mellitus (9 %) and by cardiopathology with diabetes mellitus (4 %). Ho wever, the majo rity o f patients just presented a single illness (7 4 %).

A mo ng the 2 4 7 p a tie nts w ho ha d spiro metry, 1 7 7 had no rmal pulmo nary functio n tests (7 2 %) and 7 0 patients (2 8 %) had abno rmal spiro metry, and the o bstructive pattern was the

mo st co mmo n abno rmality o bserved.

O f the 4 0 8 patients who underwent upper abdo minal surg ery, 5 8 (1 4 %) develo ped PPCs. The time spent in the intensive care unit was 6 .8 ± 6 d a ys fo r p a tie nts w ho d e ve lo p e d PPC s co mpared with the sig nificantly sho rter stay o f 2 .6 ± 2 days fo r patients witho ut PPCs (p <0 .0 5 ). Pa tie nts w ho d e ve lo p e d PPC s a lso ha d a pro lo ng ed po sto perative ho spital stay: 1 6 days (SD 7 ) fo r patients with PPCs ag ainst 8 .9 ± 6 days fo r patients witho ut PPCs (p <0 .0 5 ).

Thirty-three patients (8 %) died and in 2 2 o f these a po sto perative pulmo nary co mplicatio n was the direct cause.

The univa ria te a na lysis o f the va rio us va ria b les studied identified the fo llo wing risk facto rs that were statistically sig nificant fo r PPCs (Table 2 ): chro nic pneumo pathy (O R = 3 .6 ; 9 5 % CI = 0 .5 5 to 5 .9 0 ), presenc e o f respira to ry sympto ms at the time o f evaluatio n (O R = 2 .9 ; 9 5 % CI = 1 .7 4 to 5 .0 4 ), surg ery lasting mo re than 2 1 0 minutes (O R = 1 .8 ; 9 5 % CI = 1 .0 5 to 3 .1 2 ), co mo rbidity (O R = 1 .9 ; 9 5 % CI = 1 .1 8 to 3 .1 2 ), ag e >5 0 (O R = 1 .8 ; 9 5 % CI = 1 .0 3 to 3 .4 5 ), smo king (O R = 1 .7 ; CI = 1 .0 0 to 2 .8 9 ) and current smo king (O R = 1 .6 ; CI 1 .0 1 to 2 .6 1 ).

Table 6 - M ultivariate analysis of risk factors for postoperative pulmonary complications, in the population w ho underw ent spirometry, w ithout taking into consideration chronic pulmonary disease

Effect Estimate Standard erro r χ2 p O R (CI)

Intercept -1 .4 5 5 1 0 .1 8 7 2 6 0 .4 0 <0 .0 0 0 1 *

FEV1/ FVC <7 0 % 0 .5 1 3 3 0 .1 7 5 3 8 .5 7 0 .0 0 3 * 1 .6 6 (1 .1 9 -2 .3 1 )

Co mo rbidity -0 .3 5 5 4 0 .1 7 2 8 8 .2 3 0 .0 3 * 1 .4 1 (1 .0 2 -2 .0 6 )

DS >2 1 0 min. -0 .3 7 6 6 0 .1 8 4 7 4 .1 6 0 .0 4 * 1 .4 4 (1 .0 2 -2 .0 6 )

* Statistically sig nificant (p < 0 .0 5 ); DS: Duratio n o f surg ery; O R: O dds ratio ; CI: Co nfidence interval

Table 5 - M ultivariate analysis of risk factors for postoperative pulmonary complications, in the population w ho underw ent spirometry

Effect Estimate Standard erro r χ2 p O R (CI)

Intercept -1 .5 6 1 8 0 .2 0 3 8 5 8 .7 3 <0 .0 0 0 1 *

CPD -0 .6 8 3 5 0 .1 9 3 5 1 2 .4 8 0 .0 0 0 4 * 1 .9 7 (1 .3 6 -2 .8 6 )

DS >2 1 0 min. -0 .3 8 6 6 0 .1 7 9 3 4 .6 5 0 .0 3 * 1 .5 0 (1 .0 4 -2 .1 8 )

Co mo rbidity -0 .4 1 8 7 0 .1 9 0 6 4 .8 3 0 .0 2 * 1 .4 6 (1 .0 5 -2 .0 4 )

FEV1/ FVC <7 0 % 0 .2 6 9 8 0 .1 9 1 0 1 .9 9 0 .1 5 1 .2 9 (0 .8 9 -1 .8 8 )

(6)

N o ne the le ss, the multiva ria te lo g istic regressio n analysis identified the fo llo wing facto rs as independent predicto rs o f PPCs: presence o f chro nic pulmo nary disease, surgery lasting >2 1 0 minutes and comorbidity (Table 3 ). The concomitant o ccurrence o f all the three risk facto rs led to an even higher chance o f develo ping PPCs (Table 4 ). Amo ng st the 2 4 7 patients who perfo rmed spiro metry the incidence o f PPCs was 1 9 % (4 5 / 2 4 7 ). In this g ro up the univa ria te a na lysis ind ic a te d tha t the o c c urre nc e o f PPC s w a s asso ciated with the presence in the preo perative p e rio d o f c hro nic p ulmo na ry d ise a se (p <0 .0 0 0 1 ), respirato ry sympto ms (p = 0 .0 0 0 5 ), co mo rbidity (p = 0 .0 4 ) and current smo kers (p = 0 .0 4 ). O f the spiro metric variables analyzed, o nly an FEV1/ FVC lo wer than 7 0 % indicated a sig nificantly increase risk fo r PPCs (p = 0 .0 0 4 ). The multiple lo gistic regressio n identified the same risk facto rs in these po pulatio ns (Table 5 ).

W hen the FEV1/ FVC was co ntro lled within a multiple lo g istic reg ressio n mo del, it was no t fo und to be an independent predicto r fo r PPCs. Ho wever, further analysis revealed that if the va ria b le “ c hro nic p ulmo na ry d ise a se ” w a s withdrawn fro m the mo del, then the spiro metric value became statistically sig nificant (Table 6 ).

Using just the clinical variables it may be possible to identify the critical population. W e could calculate the pro bability o f PPCs fo r a given set o f prognostic variables using the regression coefficient and the equatio n presented in Diagram 1 .

The presence o f chro nic pulmo nary disease o r FEV1/ FVC b elo w 7 0 % identifies the sa me po pulatio n that is g o ing to develo p PPCs. This is

why FEV1/ FVC became an impo rtant predicto r o f PPCs (Table 7 ) when the same po pulatio n was a na lyz e d w itho ut ta king into a c c o unt the presence o f chro nic pulmo nary disease.

The pro bability o f PPCs o ccurring can be c a lc ula te d fro m the fa c to rs c o mo rb idity a nd surg e ry la sting mo re tha n 2 1 0 minute s, in a sso c ia tio n w ith FEV1/ FVC < 7 0 % , using equatio n 3 (Diag ram 1 )

W ith this data it was po ssible to identify eight risk classes with different chances o f develo ping PPCs in acco rdance with the prevalence o f the independent variable (Table 8 ).

DISCUSSION

Pre o p e ra tive a sse ssme nt o f p a tie nts scheduled to underg o upper abdo minal surg ery assists the physician in determining preo perative risk. Physio lo g ic a l c ha ng e s tha t o c c ur a fte r laparo to my, including alteratio ns in lung vo lume, ventilato ry gas exchange and respirato ry defense me c ha nisms, imp o se a n inc re a se d risk o f pulmo nary co mplicatio ns fo r susceptible patients. In this study we o bserved a PPC incidence o f 1 4 % (5 8 / 4 0 8 ). Ho wever, o ther studies have sho wn a wide variatio n, between 1 0 and 8 0 % in the incidence o f PPCs fo llo wing abdo minal surg ery.7 -1 0 ,2 6 ,2 7 This variatio n o ccurs because in the literature there is no standard definitio n o f PPCs a nd so the disc repa nc y b etween these finding s is understandable.

In this study the relatively lo w incidence o f pulmo nary co mplicatio ns was due to the fact that patients was underg o ing pro phylactic measures

Table 7 - M ultivariate analysis of clinical risk factors for postoperative pulmonary complications in the population w ho underw ent spirometry

Effect Estimate Standard erro r χ2 p O R (9 5 % CI)

Intercept -1 .6 5 1 8 0 .1 9 7 0 7 0 .2 9 < 0 .0 0 0 1 *

CPD -0 .7 7 3 3 0 .1 8 2 8 1 7 .9 0 <0 .0 0 0 1 * 2 .1 6 (1 .5 2 -3 .0 6 )

Co mo rbidity -0 .4 1 1 5 0 .1 7 8 0 5 .3 4 0 .0 2 0 * 1 .5 0 (1 .0 8 -2 .1 0 )

DS >2 1 0 min. -0 .4 1 8 4 0 .1 9 0 1 4 .8 4 0 .0 2 7 * 1 .5 0 (1 .0 4 -2 .1 8 )

* Statistically sig nificant (p < 0 .0 5 ); CPD: Chro nic pulmo nary disease; DS: Duratio n o f surg ery; O R: O dds ratio ; CI: Co nfidence

(7)

to avo id such co mplicatio ns, as reco mmended by the ethical co mmittee.

Bro ncho co nstrictio n was the mo st frequent co mplicatio n o bserved (5 0 % = 2 9 / 5 8 ). It was no t asso ciated with intubatio n o r extubatio n and usua lly o c c urre d a lo ne . M o re tha n o ne co mplicatio n is frequently o bserved in the same patient at the same time.

Pulmo na ry infe c tio n (p ne umo nia o r tracheo bro nchitis) acco unted fo r 4 0 percent (2 3 / 5 8 ) o f the patients with PPCs. Sixty-five percent o f the se p a tie nts (1 5 / 2 3 ) d e ve lo p e d a c ute respirato ry failure with pro lo ng ed intensive care treatment and ho spitalizatio n. The hig h ho spital co sts impo sed by this additio nal medical care have been described in the literature.2 8

This study was designed to evaluate patients underg o ing elective upper abdo minal surg ery, using c linic a l a nd spiro metric pa ra meters a s sug g ested by previo us studies, so as to estimate the pro bability o f PPCs. Befo re do ing this study, we used to assess the pro g no stic index o n the basis o f pro g no stic sco res develo ped in o ther po pulatio ns with different backg ro unds. As well a s eng endering a sig nific a nt impro vement in pro g no sis, risk facto rs co uld be calculated based o n stratified risk g ro ups that we develo ped in this wo rk, unlike in mo st o ther studies (Table 8 ). The main facto rs traditio nally asso ciated w ith PPC s a re c hro nic a irw a y s d ise a se , a d va nc e d a g e , up p e r a b d o mina l surg e ry, p ro lo ng e d d ura tio n o f surg e ry, histo ry o f smo king , and o besity.1 ,7 ,8 ,1 0 ,1 3 ,2 8 ,2 9

To select the risk facto rs to be included in lo g istic reg ressio n fo r PPC we cho se the variables based o n p value o f the chi square test in a step-do wn sig nificance pro cedure. The variables were included in the fo llo wing sequence: pneumo pathy (p = 0 . 0 0 0 1 ), re sp ira to ry symp to ms (p = 0 .0 0 0 1 ), co mo rbidity (p = 0 .0 0 7 ), ag e o ver 5 0 years o ld (p = 0 .0 3 ), surg ery duratio n o ver 2 1 0 minutes (p = 0 .0 3 ), habitual smo king (p = 0 .0 4 ), c urre nt smo king (p = 0 . 0 4 ) a nd d ystro p hy (p=0 .1 0 ) (Table 2 ).

The lo g istic reg ressio n in o ur study sho wed the same risk facto rs as fo und in the literature. W e fo und that the presence o f chro nic pulmo nary

disease is also stro ng ly asso ciated with PPCs, particularly if the patient’s surg ery lasted lo ng er than 2 1 0 minutes and co mo rbidity was present. Chro nic pulmo nary disease is o ne o f the mo st c o mmo n risk fa c to rs fo und in the lite ra ture .8 ,2 6 ,2 9 ,3 0 -3 2 O nc e c hro nic p ulmo na ry d ise a se ha s b e e n ta ke n into a c c o unt, the presence o f respirato ry sympto ms and smo king habits ceases to be stro ng ly asso ciated with a hig h incidence o f PPCs. Mitchell et al,1 1 in a study o f 2 0 0 patients who underwent g eneral surg e ry, o b se rve d tha t the c o ntrib utio n o f cig arette smo king to PPCs is mo re likely to o ccur via its asso ciatio n with hypersecretio n o f mucus.

W illia ms-Russo et a l,9 in a study o f the p re d ic tive va lue o f c o mo rb id ity fo r PPC s, c o nc lud e d tha t p re d ic ting a nd p re ve nting po sto perative cardiac mo rbidity may be the best appro ach fo r reducing po st-o perative pulmo nary mo rbidity. The duratio n o f surg ery is also an impo rtant risk facto r fo r PPCs.9 ,1 1 ,1 4

Thus, when o ne patient presents the three c o nc o mita nt risk fa c to rs (c hro nic p ulmo na ry disease, surg ery lasting o ver 2 1 0 minutes and co mo rbidity), the pro bability o f his develo ping a po sto perative pulmo nary co mplicatio n increase to 4 7 %. The sensitivity and specificity fo r this

Table 8 - Relationship betw een presence of risk factors and predictability of postoperative respiratory complication (Equation 1 for previous respiratory disease and Equation 3 for FEV1/ FVC)

Risk %

CPD or FEV1/ FVC <70% DS > 210 Comorbidity Equation 1-3

N o N o N o 1 4 1 9

N o N o Yes 2 0 2 5

N o Yes N o 2 1 2 5

Yes N o N o 2 7 2 8

N o Yes Yes 2 8 3 2

Yes N o Yes 3 6 3 5

Yes Yes N o 3 7 3 6

Yes Yes Yes 4 7 4 4

CPD: Chro nic pulmo nary disease; DS: Duratio n o f surg ery >

(8)

test were 9 2 .8 and 7 9 .4 respectively (Table 4 ), and the po sitive predictive value was 4 4 .8 %. This allo wed us to identify the critical po pulatio n that needed intensive preo perative treatment.

The se find ing s ha ve ma ny c linic a l implicatio ns, because all hig h-risk patients can b e id e ntifie d b y the p re se nc e o f p re vio us respira to ry disea se, c o mo rb idity a nd surg ery d ura tio n o f mo re tha n 2 1 0 minute s. W he n po ssible these facto rs sho uld be mo dified in o rder to reduce the incidence o f PPCs.

Ro utine preo perative spiro metry fo r patients who are to have upper abdo minal surg ery is reco mmended to identify and beg in preventive care fo r tho se with abno rmal finding s.8 ,3 3 ,3 4

Spiro metry co uld no t be perfo rmed o n all the individuals in this study. Ho wever, it was indicated fo r patients in acco rdance with the criteria o f Ho usto n et al,8 which take into acco unt

no t o nly patients with pulmo nary pro blems and respirato ry sympto ms, but also all patients with ag es equal o r superio r to 6 5 years o ld, and all pa tie nts e xpo se d to smo king inde pe nde nt o f respirato ry sympto ms o r age. This is different fro m the criteria established by Zibrak et al3 5 fo r pre-o perative evaluatipre-o n, which pre-o nly include thpre-o se patients with abno rmal pulmo nary histo ry fo r do ing spiro metry. The Ho usto n criteria allo wed us to g ua ra nte e ha ving a re p re se nta tive po pulatio n similar to the entire g ro up o f patients by admitting into the subpo pulatio n so me patients with a no rmal pulmo nary functio n test. O n the o the r ha nd , the c o nfirma tio n tha t the subpo pulatio n had the same characteristics as the entire g ro up pro vided us with a result that sho uld p ut in d o ub t the re a l va lid ity o f the ind ic a tio n fo r sp iro me try re c o mme nd e d b y Ho usto n et al.

Thus, the sample was representative o f the g ro up and included patients with no rmal and abno rmal functio ning . Furthermo re, in o rder to e nsure tha t w e w e re w o rking w ith a subpo pulatio n (2 4 7 patients) representative o f the who le g ro up, we o nce ag ain carried o ut a lo g istic reg ressio n a na lysis o f the risk fa cto rs id e ntifie d b y univa ria te a na ly sis o n this subpo pulatio n and we o btained the same result in relatio n to the to tal o f patients (Equatio n 2 ).

In o ur po pulatio n, pulmo nary functio n tests do a ppe a r to ha ve b e e n o f so me b e ne fit in predicting PPCs, especially an FEV1/ FVC lo wer than 7 0 %. Ho wever, when we used multivariate a na lysis to d e te rmine the imp a c t o f e a c h independent va ria b le o n the o utc o me, in the co ntext o f all o ther variables, we co uld see that the presence o f altered FEV1/ FVC appeared to be sig nificantly asso ciated with the o ccurrence o f PPCs (Table 6 ). This o ccurs because the FEV1/ FVC alteratio n can no t be disso ciated fro m the presence o f chro nic pulmo nary disease. In o ther w o rd s, in p a tie nts w ith c hro nic p ulmo na ry disease, the measurement o f FEV1/ FVC wo uld o nly be helpful if we were to g ive up do ing a g o o d clinical evaluatio n.

Lawrence et al33 assessed the predictive value o f preo perative spiro metry thro ugh a systematic

Diagram 1 - Equations for estimating the probability of post operative pulmonary complications in patients submitted to elective

upper abdominal surgery, using logistic regression. General population: Equation 1. Spirometric population: Equation 2 for clinical variables and Equation 3 for spirometric variables Equation 1:

1

P(y)=

---1 + e-(-1 .7 9 + 0 .8 1 CPD + 0 .4 6 DS + 0 .4 0 C)

Equation 2:

1

P(y)=

---1 + e-(-1 .6 5 + 0 .7 7 CPD + 0 .4 1 C + 0 .4 2 DS)

Equation 3:

1

P(y)=

1 + e-(-1 .4 6 + 0 .5 1 FEV1/ FVC + 0 .3 5 C + 0 .3 8 DS)

P(y): Pro bability o f po sto perative pulmo nary co mplicatio n;

CPD: Chro nic pulmo nary disease; DS: Duratio n o f surg ery >

(9)

litera ture sea rc h a nd c ritic a l a ppra isa l o f the published literature and concluded that it is not clear that spirometry adds much predictive value beyond that of a clinical examination alone, and that the full po tential o f spiro metry fo r precise, accurate risk assessment may not yet have been realized.

Spiro metry can o nly be reco mmended when patients being prepared fo r abdo minal surgery are cigarette smo kers o r have respirato ry co mplaints that have no t been previo usly evaluated.3 4

O ur d a ta sug g e sts tha t w he n g e ne ra l preo pera tive eva lua tio n do es no t revea l a ny c la ssic histo ry o f lung d ise a se , p ulmo na ry functio n testing may assist in making a specific pulmo nary diag no sis and assessing the deg ree o f impairment befo re o perating .

This study co nfirms the impo rtance o f an evaluatio n o f a patient’s respirato ry co nditio n, e sp e c ia lly a c linic a l e va lua tio n, in o rd e r to determine the risk fo r PPCs in a g iven po pulatio n. Thus it is difficult to avo id the impressio n that the best predicto r o f the o verall o perative risk in the individual patient is still a co mprehensive c linic a l e va lua tio n. Pre dic tio n rule s a nd risk stratificatio n sho uld be metho do lo gically so und, c linic a lly va lid a te d a nd ho p e fully w id e ly accepted.3 4

REFERENCES

1. Mo rto n AP. Respirato ry preparatio n fo r abdo minal surgery. Med J

Aust 1973;1:1300-4

2. Dureuil B, Cantineau JP, Desmo nts JM. Effects o f upper o r lo wer abdo

mi-nal surgery o n diaphragmatic functio n. Br J Anaesth 1987; 59:1230-5.

3. Dureuil B, Cantineau JP, Vo gel J, Desmo nts JM. Vital capacity and

diaphragmatic functio n after abdo minal surgery. Anesthesio lo gy 1984;61:A478.

4. Fo rd GT, Guenter CA. To ward preventio n o f po sto perative pulmo

-nary co mplicatio ns. Am Rev Respir Dis 1984;130:4-5.

5. Fo rd GT, Whitelaw WA, Ro senal TW, Cruse PJ, Guenter CA. Diaphragm

functio n after upper abdo minal surgery in humans. Am Rev Respir Dis 1983;7:431-6.

6. Fo rd GT, Ro senal TW, Clergue F, Whitelaw WA. Respirato ry physio lo gy

in upper abdo minal surgery. Clin Chest Med 1993;14:237-52.

7. Jackso n CV. Preo perative pulmo nary evaluatio n. Arch Intern Med

1988;148:2120-6.

8. Ho usto n MC, Ratcliff DG, Hays JT, Gluck FW. Preo perative medical

c o n s u ltatio n an d e valu atio n o f s u rg ic al ris k. So u th Me d J 1987;80:1385-96.

9. Williams-Russo P, Charlso n ME, Mackenzie R, Go ld P. Predicting

p o s to p e rative p u lm o n ary c o m p lic atio n s . Arc h In te rn Me d 1992;152:1209-13.

10. Hall JC, Tarala A, Hall JL, Mander J. A multivariate analysis o f the risk

o f pulmo nary co mplicatio ns after laparo to my. Chest 1991;99:923-7.

11. Mitchell C, Garrahy P, Peake P. Po sto perative respirato ry mo rbidity:

identificatio n and risk facto rs. Aust NZ J Surg 1992;52:203-9.

12. To rringto n KG, Henderso n CJ. Preo perative respirato ry therapy. A

pro gram o f preo perative risk assessment and individualized po st-o perative care. Chest 1988;93:946-51.

13. Celli BR. Preo perative respirato ry care o f the patient undergo ing

upper abdo minal surgery. Clin Chest Med 1993;14:253-62.

14. Latimer RG, Dickman M, Day WC, Gunn ML, Schmidt CD. Ventilato ry

patterns and pulmo nary co mplicatio ns after abdo minal surgery deter-mined by preo perative and po sto perative co mputerized spiro metry and blo o d gas analysis. Am J Surg 1971;122:622-32.

15. American Tho racic So ciety Statement. Standards fo r the diagno sis

and care o f patients with chro nic o bstructive pulmo nary disease. Respir Crit Care Med 1995;152(supp):77-120.

16. NHLBI. Natio nal asthma educatio n pro gram, expert panel repo rt:

Guidelines fo r diagno sis and management o f asthma. J Allergy Clin Immuno l 1991;88:425-534.

17. NHLBI. Internatio nal co nsensus repo rt o n diagno sis and

manage-ment o f asthma. Bethesda, MD: Natio nal Institutes o f Health publ; 1992:92-3091.

18. Frankel HM. Determinatio n o f bo dy mass index. JAMA 1986;255:1292.

19. NHI. Natio nal ins titute s o f he alth c o ns e ns us d e ve lo p m e nt

co nference statement. Health implicatio ns o f o besity. Ann Intern Med 1985;103:147-51.

20. American Thoracic Society. Lung function testing: selection of reference

values and interpretational strategies. Am Rev Respir Dis 1991;144:1202-8.

21. Celli BR, Ro driguez KS, Snider GL. A co ntro lled trial o f intermittent

po sitive pressure breathing, incentive spiro metry and deep breathing exercises in preventing pulmo nary co mplicatio ns after abdo minal surgery. Am Rev Respir Dis 1984;130:12-15.

22. Murphy TF, Sethi S. Bacterial infectio n in chro nic o bstructive pulmo nary

disease: state o f the art. Am Rev Respir Dis 1992;146:1067-83.

23. So x HC. Pro b ab ility the o ry in the use o f d iagno stic te sts. An

intro ductio n to critical study o f the lite rature . Ann Inte rn Me d 1986;104:60-6.

24. Ranso ho ff DF, Feinstein AR. Pro blems spectrum and bias in evaluating

the efficacy o f diagno stic tests. N Engl J Med 1978;26:926-9.

25. Co ncato J, Feinstein AR, Ho lfo rd TR. The risk o f determining risk

with multivariable mo dels. Ann Intern Med 1993;118:201-10.

26. Dureuil B, Vires N, Cantineau JP, Leco cguic J, Marty C, Aubier M,

De sm o nts JM. Me c hanism o f d iap hragm atic d ysfunc tio n afte r abdo minal surgery. Anesthesio lo gy 1984;61:A479.

27. Co llins CD, Darke CS, Kno welden J. Chest co mplicatio ns after upper

abdominal surgery: their anticipation and prevention. Br Med J 1968;1:401-6.

28. Olsso n GL. Bro ncho spasm during anaesthesia. A co mputer-aided

inc id e nc e stud y o f 136929 p atie nts. Ac ta Anae sthe sio l Sc and 1987;31:244-52.

29. Martin LF, Asher EF, Casey JM, Fry DE. Po sto perative pneumo nia.

Arch Surg 1984;119:379-83.

30. Kro enke K, Lawrence VA, Thero ux JF, Tuley MR. Operative risk in

patients with severe o bstructive pulmo nary disease. Arch Intern Med 1992;152:967-71.

31. Milledge JS, Nunn JF. Criteria o f fitness fo r anaesthesia in patients

with chro nic o bstructive lung disease. Br Med J 1975;3:670-3.

32. Tisi GM. Preo perative evaluatio n o f pulmo nary functio n. Am Rev

Respir Dis 1979;119:293-310.

33. Lawrence VA, Page CP, Harris GD. Preo perative spiro metry befo re

abdo minal o peratio ns: a critical appraisal o f its predictive value. Arch Intern Med 1989;149:280-5.

34. Zibrak JD, O’Do nnell CR, Marto n K. Indicatio ns fo r pulmo nary

(10)

35. Hirshberg A, Adar R. Preo perative predictio n o f po sto perative co m-plicatio ns. Isr J Med Sci 1990;26:123-4.

Ea nes Delga do Ba rros Pereira - MD, PhD, Asso ciate

Pro fesso r, Department o f Medicine, Pulmo nary Divisio n, Federal University o f Ceará - Brazil

Ana Luisa Godoy Ferna ndes - MD, PhD, Asso ciate

Pro fesso r, Department o f Medicine, Pulmo nary Divisio n, Federal University o f São Paulo - Brazil

M eide da Silva Ançã o - MD, PhD, Asso ciate Pro fesso r,

Department o f Medicine, Divisio n o f N ephro lo g y, Federal University o f São Paulo - Brazil

Clóvis de Ara újo Peres - PhD, Pro fesso r o f Statistics o f

University o f São Paulo

Álvaro N agib Atallah - MD, PhD, Head o f Internal

Medicine Divisio n o f Federal University o f São Paulo - Brazil.

RESUMO

Objetivo: Avaliar asso ciaçõ es entre a presença de variáveis pré-o perató rias e a o co rrência de co mplicaçõ es pulmo nares no

pó s-o perató rio de cirurg ia abdo minal alta eletiva. Tipo de Estudo: Experimentação clínica em perspectiva. Pa cientes: 4 0 8 pacientes fo ram avaliado s pro spectivamente no pré-o perató rio e seg uido s no pó s-o perató rio para avaliação de co mplicaçõ es pulmo nares. Va riá veis medida s: Características do s pacientes: histo ria clínica e exame físico , do enças asso ciadas, tabag ismo , e tempo cirúrg ico . Fo i realizado espiro metria no perío do pré-o perató rio em 2 4 7 pacientes. Resulta dos : A incidência de co mplicaçõ es pulmo nares no pó s-o perató rio fo i 1 4 %. N a análise univariada co mpo rtaram-se co mo fato res de risco para o co rrência de co mplicaçõ es pulmo nares no pó s-o perató rio : idade acima de 5 0 ano s, tabag ismo , a presença de do ença pulmo nar crô nica o u sinto mas respirató rio s no mo mento da avaliação , tempo cirúrg ico maio r que 2 1 0 minuto s e a co existência de do ença clínica (p< 0 ,0 4 ). N a análise de reg ressão lo g ística co mpo rtaram-se co mo fato res de risco so mente a presença de do ença pulmo nar crô nica, tempo cirúrg ico maio r que 2 1 0 minuto s e co existência de do ença clínica asso ciada. (p< 0 ,0 0 9 ). Conclusã o: Existem três fato res de risco clínico s para co mplicaçõ es pulmo nares no pó s-o perató rio de cirurg ia abdo minal alta: do ença pulmo nar crô nica, tempo cirúrg ico maio r que 2 1 0 minuto s e co existência de do ença clínica. O s pacientes que apresentam estes fato res de risco têm três vezes mais chance de co mplicar do po nto de vista pulmo nar no pó s-o perató ris-o quands-o cs-o mparads-o s aqueles que nãs-o apresentam estes fats-o res de riscs-o (p< 0 ,0 0 1 ). A espirs-o metria está indicada quando não existe certeza quanto ao quadro pulmo nar do paciente.

Sonia M a ria Fa resin - MD, PhD, Physician o f

Department o f Medicine, Pulmo nary Divisio n, Federal University o f São Paulo - Brazil. Chief o f Preo perative Respirato ry Care o f São Paulo Ho spital

Sources of Funding: Partially suppo rted by CN Pq grant

Conflict of interest: N o t declared

La st received: 8 February 1 9 9 9

Accepted: 2 March 1 9 9 9

Address for correspondence:

So nia Faresin

Universidade Federal de São Paulo - Disciplina de Pneumologia Rua Bo tucatu, 7 4 0 - 3 º andar

Referências

Documentos relacionados

OBJECTIVE : To describe the clinicopathological characteristics of patients with upper urinary tract transitional cell carcinomas who are treated surgically and to analyze

analyzed the complications that increase the permanence of the patients submitted to cardiac surgery at intensive care unit and conclude that they are related to respiratory

The objective of the present study was to identify the independent variables associated with postoperative pulmonary complications in sarcoma patients undergoing

The objective of this study was to compare in a prospective randomized study the immediate postoperative clinical results between the two groups of patients who were submitted

Prediction of postoperative pulmonary complications on the basis of preoperative risk factors in patients who had undergone coronary artery bypass graft surgery. American

Feasibility of preoperative inspiratory muscle training in patients undergoing coronary artery bypass surgery with a high risk of postoperative pulmonary complications: a

In nonobese patients submitted to upper abdominal surgery (UAS), changes in respiratory mechanics, respiratory pattern, gas exchanges and lung defense mechanisms occur during

The objective of the present study was to identify the independent variables associated with postoperative pulmonary complications in sarcoma patients undergoing