Registro de Complicaciones

7/28/2019 Registro de Complicaciones

1/6

The Accuracy of Complications Documented in a Prospective

Complication Registry

Eelco J. Veen, M.D., Ph.D.,*,1 Maryska L. G. Janssen-Heijnen, Ph.D., Eelke Bosma, M.D.,Maryska A. C. de Jongh, and Jan A. Roukema, M.D., Ph.D.

*Department of Surgery, Amphia Hospital, Breda, The Netherlands; Comprehensive Cancer Centre South, Eindhoven,The Netherlands; and Department of Surgery, St. Elisabeth Hospital, Tilburg, The Netherlands

Originally submitted January 10, 2010; accepted for publication August 23, 2010

Background. The objectives of this study were to

evaluate the accuracy of a prospective complication

registry for documenting complications and identify

possible factors for non-registering.Methods. Five hundred randomly selected patients

admitted at the Department of Surgery of St. Elisabeth

Hospital Tilburg, The Netherlands, in the year 2005,

were evaluated for incidence and type of complica-

tions by an examination of their medical records and

compared with a prospective complication registry.

The system was independently reviewed by two per-

sons for missing complications. Patient files with miss-

ing complications in the registry were screened for

factors possibly responsible for non-registering.

Results. Two hundred thirteen complications were

detected, 58 (27%) missing in the registry. There were

50 different types of complications documented. The

number of events missing per category were: drug-

related (50%, n [ 4), organ dysfunction (44%, n [ 14),

infection-related (25%, n [ 19), surgery/intervention-

related (23%, n [ 14), and hospital-provider errors

(19%, n[ 7). Not all clinically important complications

were adequately documented (e.g., anastomotic leak-

age). The kappa score was 0.695, making the interrater

reliability substantial.

Conclusion. The accuracy of registering complica-

tions is fairly acceptable compared to the ranges men-

tioned in literature. It is disappointing that clinically

important events are missing in the registry. The inac-

curacy could be explained by a great diversity of docu-

mented events, due to a broad definition, suggestingignorance of the responsible team of which events to

register. Crown Copyright 2012 Published by Elsevier Inc. All rights

reserved.

KeyWords: complication registration; quality assess-

ment; outcome measurement.

INTRODUCTION

Clinical audit is the systematic and critical analysisof quality of care, in which emphasis is put on improve-ment [1, 2]. In surgery, the recording of complications isused as an instrument for audit, with great variabilityin the registry process [35]. The success of an auditprogram is very much dependent on the quality of thedata collected; this creates the need to obtainaccurate, standardized information on activities andoutcome [6]. The purpose of a database is to provide in-formation that could be used as a benchmarking tool forquality assurance and improvement. Poor informationfrom a database results in bias and unreliable outcome[7]. Whether and to what extent differences in outcomesreflect differences in quality of care is not clear. In par-ticular, outcome measures must be adjusted for casemix, such as age, co-morbidity, underlying disease,and type of surgery. Furthermore, there are importantpractical problems involved in using complications forquality measurement, such as definition development,the sources of data on complications, and their reliabil-ity [8]. Consequently, like any other measurementtools, those used for adverse events need to be tested

for their accuracy. We have been prospectively register-ing complications in our clinic with the intention ofusing the register as an instrument for quality assur-ance. The objectives of this study were to evaluate theaccuracy of our registry system by evaluating typeand incidence of complications recorded and factorspossibly responsible for missing complications.

1 To whom correspondence and reprint requests should be ad-dressed at Department of Surgery, Amphia Hospital, Molengracht21, 4818 CK Breda, The Netherlands. E-mail: [email protected].

0022-4804/$36.00Crown Copyright 2012 Published by Elsevier Inc. All rights reserved.

54

Journal of Surgical Research 173, 5459 (2012)doi:10.1016/j.jss.2010.08.042
mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.jss.2010.08.042http://dx.doi.org/10.1016/j.jss.2010.08.042http://dx.doi.org/10.1016/j.jss.2010.08.042http://dx.doi.org/10.1016/j.jss.2010.08.042mailto:[email protected]://dx.doi.org/10.1016/j.jss.2010.08.042http://dx.doi.org/10.1016/j.jss.2010.08.042


2/6

PATIENTS AND METHODS

Patients and Validation

The method for validating our registry system was performed byevaluating a randomly selected group of 500 (11%) patients admittedat our surgical department in the year 2005. Each case recordwas ex-aminedthoroughlyfor incidenceand type of complications, by thefirst

author (EJV), who was blinded for the complications documented inthe registry. The examination was performed according to the defini-tion of a complication developed by the Association of Surgeons of theNetherlands (ASN): Any state or event, unfavorable to the patientshealth, which arose during admission or within 30 d after dischargethat either causes unintentional injury or requires additional treat-ment. Next, of 495 eligible records, a random sample of 100 recordswas drawn, using SPSS 16.0, SPSS Inc., Chicago, IL. All randomlydrawn records were fully reviewed for missing complications by a sec-ond reviewer (EB). It was scored on whether a patient was consideredto have experienced complications, and the medical record wasreviewed for complications that were not recorded in theregistry. The follow-up period was chosen to be 30 d according tothis definition. A patient centered approach is used, meaningthat the whole care process is evaluated. Patient confidentiallywas maintained. Complications were categorised as: (1) surgery/

intervention-related, (2) infection-related, (3) organ dysfunction (car-diopulmonaryrenalneurologicalgastrointestinal), (4) drug re-lated, and (5) hospital-provider errors, divided into the followingevents: incomplete hospital record, delay in disposition, delay in MDresponse, delay in obtaining consultation, delay in diagnosis, errorin diagnosis, error in judgement, error in technique, andother miscel-laneous. The electronic file contains, in addition to all patients andtreatment characteristics, the report of the procedure performed, alldischarge letters (for instance from the intensive care unit), labora-tory values, pathology exams, and diagnostics performed. Patientswith missing complications were compared with the group of patientswithout missing events. The followingcharacteristics wereevaluated;age, diagnosis (general surgery, vascular surgery, traumatology, on-cology, gastro-intestinal surgery), [American Society of Anaesthesiol-ogists (ASA) Physical Status classification [9], gender, procedure (yes/

no andacuteor elective), complexity of surgery; classified accordingtothe year in which a surgical trainee is expected to be able to performthe operation, defined by the Association of Surgery of the Nether-lands [10], and duration of hospital stay.

Complication Registry

The registration methods and classifying systems used have beendescribed in detail elsewhere [11, 12]. In short, complications areprospectively recorded in our registry, which forms part of theelectronic medical patient file. A complication is identified realtime and recorded by one of the physicians in our surgical team(12 surgical trainees and nine consultants). The electronicmedical file is operational all over the hospital and at the out-patient clinic. In-hospital and at the out-patient clinic, documentedevents are automatically presented at the daily surgical conferenceand discussed for coding and type of surgical complication by thesurgical team. The complication is classified according to the sys-tems of the Association of Surgery of The Netherlands (ASN) andthe Trauma Registry of the American College of Surgeons (TRACS).Neither system gives information about severity. The system of the

ASN uses four denominators to classify an event: nature of thecomplication, anatomic localization, specification, and additional de-scription. The TRACS was originally developed as a complicationlist to record the morbidity in trauma patient populations. Thelist explicitly defines complications and uses four-digit-codes. Al-though this list was developed for the trauma population, itsdesign is broad and encompasses complications applicable to gen-eral surgery. Finally, a description of the complication is recordedin full text.

Statistical Analysis

Statistical analyses were performed through a computerizedsoftware package using Excel (Office XP form Microsoft) and Graph-Pad Prism 4. The analysis of registered events of the two groups inthe registry and missing events was performed with the unpairedindependent Students t-test, c2 statistic, and Mann-Whitney testfor nonparametric data. An inter-rater reliability analysis using thek statistic was performed to determine consistency between both re-viewers. Values of k from 0.40 to 0.59 were considered moderate,0.60 to 0.79 substantial, and 0.80 outstanding (Landis andKoch, 1977).

RESULTS

Non-Registered Complications

Five patients were excluded, as they were admitted toanother specialty than surgery; general characteristicsof the 495 patients analyzed are shown in Table 1. Hos-pital stay and age significantly differed between the

groups with and without complications. After screeningall patient files, the auditor detected 213 complications,58 (27%) were missing in the registry. The analysis of100 selected patients by the second reviewer showedmissing complications in 13 patients. In eight patientsthe reviewers agreed to missing events, with a substan-tial inter-rater reliability, k score was 0.695. Table 2shows the number and proportions of registered andnon-registered events, according to the category ofthe complication. In the category organ dysfunction(n 32), 13 different types of complications were re-corded, with 14 (44%) complications missing, dividedover seven types of events. Ten out of 14 were

TABLE 1

Characteristics of 495 Admitted Patients at the

Department of Surgery of the Elisabeth Hospital

Tilburg in the Year 2005

DiagnosisGeneral surgery n 219 (44%)

Vascular surgery n 72 (15%)Traumatology n 88 (18%)Oncology n 61 (12%)Gastro-intestinal n 55 (11%)

Age (median y range) 53 (091)

Hospital admission (mean d range) 5 (1109)Patients with complications n 144 (29%)

Age (median y range) 59 (891)*Hospital stay (mean d range) 10 (1109)*

Patients without complications n 351 (71%)Age (median y range) 49 (091)Hospital stay (mean d range) 5 (136)

GenderMale/female 258 (52%)/237 (48%)

Procedures:Acute n 107 (21%)Elective n 311 (63%)No Procedure n 77 (16%)

*P < 0.001.

VEEN ET AL.: ACCURACY OF DOCUMENTED COMPLICATIONS IN SURGERY 55


3/6

pulmonary-related: respiratory failure (n 7) in fourpatients, acquired respiratory distress syndrome intwo patients, and one pulmonary embolus. In contrast,all cardiac events (myocardial infarction: n 4/ar-rhythmia: n 5/congestive heart failure n 3) were ad-equately registered. In the surgery/interventionrelated category (n 56) half of the postoperative oc-cluded vascular grafts (n 10) (three arterial-venousfistula, one percutaneous transluminal angioplastywith stent placement of the iliac artery, and one femoro-popliteal bypass) were missing in the registry. Postop-erative hemorrhage was not documented in one (6%)patient. Clinically important complications such as iat-rogenic pneumothorax (occurring at the ICU and recov-ery room) and anastomotic leakage were both foundtwice in the patient file, yet registered only once. Onebile leak after laparoscopic cholecystectomy was notedin the patient file; however, it was not documented.

This patient was re-admitted 5 d after the initial proce-dure and was then diagnosed with common bile ductstones. An endoscopic retrograde cholangiopancreati-cography (ERCP) was performed, which also showeda bile leak form the cystic duct. Among infection-related complications, pneumonia and wound infectionwere not registered in 38% and 20%, respectively. Inthe drug-related category, the four missing eventswere two anticoagulant-related bleedings and two al-lergic reactions.

Type of Events

Fifty different types of events were recorded. The cat-egory hospital-provider (n 36) (Table 3) appeared tobe most diverse, with 27 different types of events dis-tributed over seven categories.

Events Only Documented in the Registry

Thirty-seven complications were only documented inthe registration system and not detected by the auditor.These events were divided over three categories, re-spectively, hospital-provider errors (n 26) with 20

different types of complications, drug-related (n 2),and nine complications that could not be categorised.Evaluating these events, we concluded that 20 (54%)did not meet the criteria of a complication as definedabove. For instance, a cicatrical hernia after a laparo-scopic gynecologic procedure a few years before, recur-rent varicosis 9 y after the first procedure, a patient

with Crohns disease presenting with intra-abdominalabscess formation, irradical excision of breast carci-noma, mononeuritis simplex in a patient with SLE,postoperative abdominal pain after a cholecystectomy.

Characteristics of Patients with Complete and

Incomplete Registry of Complications

There were 144 patients with complications; in 88 pa-tients, the registry was complete in documenting com-plications, and in 42 patients, events were missing,the other 14 were excluded as this consisted of patients

with complications only documented in the registry. InTable 4, the differences between the two groups areshown. Significantly more complications per patientin the group with missing events were noticed. Therewere fewer patients with missing complications inthe category general surgery and relatively more inthe traumatology population. In patients with missingcomplications, significantly more acute and complexprocedures were performed, and their hospital staywas twice as long compared with patients with ade-quate recording of complications.

DISCUSSION

The intention of registering complications is to pro-duce information about quality of care, which shouldlead to health care improvement activities. As in otherregistries, data must be of good quality and should betested for potential errors, to be useful in this aspect.In this study, our complication registry was evaluatedfor accuracy in type and incidence of documenting com-plications. In this study, 213 complications were foundin medical files of 495 consecutive patients, of which 58(27%) events were missing in the departments prospec-

tive complication registry. To find the true incidence ofcomplications occurring in surgery, a representativebenchmark should be selected to compare data. Retro-spective medical record review was developed by theHarvard Medical Practice Study (HMPS) in 1984 andwas proven to be valid in identifying adverse events[13]. Our methods are also retrospective, but not vali-dated as the Harvard Medical Practice Study, whichcould be a flaw of our study. Therefore, a second re-viewer additionally evaluated 100 patient files for miss-ing complications, with a substantial inter-ratervariability, k score of 0.695. As a single center study,

TABLE 2

Proportion of Missing Complications and According to

the Category of Complications

Category ofcomplications

Complicationsafter review

(n 213)

Missingcomplications in registry

(N58/27%)

Surgical/intervention n 56 (26%) n 14 (25%)Infection n 81 (38%) n 19 (23%)Organ related n 32 (15%) n 14 (44%)Drug related n 8 (4%) n 4 (50%)Hospital-provider n 36 (17%) n 7 (20%)

JOURNAL OF SURGICAL RESEARCH: VOL. 173, NO. 1, MARCH 201256


4/6

these results show that accurate data gathering is pos-sible, however, there is still room for improvement. De-fining the accuracy of a database according to anacceptable rate of missing events is difficult. Numbersof missing data were 25% in the study on the UK car-

diac surgery experience [14]. A study by Bertges et al.,evaluating the recording of complications in vascularsurgery measured by two different systems, showedmissing events in the in hospital database of 29%, and18% in the hospital discharge data set [15].

TABLE 3

Type and Number of Hospital-Provider Errors and Missing Events

Hospital-provider error n 29 Missing events per type of provider error (n 7)

Other miscellaneous n 1Dislocated pigtail tube 1 No capable surgeon, procedure canceledPatient fallen out of bed, head injury 1

Delay in MD response n 1Incomplete or incorrect preoperative workup 5 sutures after several weeks still in situDecubitus (deep) 1

Delay in diagnosis n 1Cervical spine fracture primary missed 1 Delay in diagnosing thoracic spine fracture

Error in judgement n 1Deep infection not drained in the OR 1 Medication given to wrong patient.Wrong interpretation of the electrocardiogram 1Too high dose nitroglycerine given 1Not all cortical screws removed 1Wrong perianal pathology treated 1Morphine dose is 10-fold too high respiratory failure after detubation 1

Error in technique n 3Echo guided wire for breast tumor inaccurately placed 1 Intramedullary nail too longStent dislocated and lost with endovascular procedure 2 Thorax tube-intrapulmonary

Not all varicosis treated as preoperative discussed 1 Two extra cortical screw necessary and placedtwo days after primary procedureSubphrenic abscess drained through the pleura 1Duodenum tube localised in the stomach 1No hemoclips left behind after lumpectomy, violation of the protocol 1

TABLE 4

Characteristics of Patients with Complete and Incomplete Number of Registered Complications*

Patients with registeredcomplicationsn 88/70%

Patients with missingcomplicationsn 42/32% P value

No. of complications in patient file n 114 n 99complications/patient n 1 n 3


5/6

In our study we missed five occluded vascular grafts;three were arterial-venous fistula for hemodialysis.Those patients left the same day and the nephrologydepartment performed follow-up, which could be anexplanation for missing/non-registering of theseocclusions. Proportion of missed events in a study byGunnarson et al. in colorectal cancer surgery differed

between 0.06 and 0.69 among three databases evalu-ated [16]. In a Danish national database on case regis-try of laparoscopic cholecystectomy, completeness ofregistration was 69%, 80%, and 99%, respectively,among three randomly chosen departments [17]. Ac-cording to the above-mentioned studies, the proportionof missing events in our study seems to be fairly accept-able, although comparison is difficult due to differencein recording methodologies.

However, some clinically important complications(e.g., anastomotic leakage, bile leakage, iatrogenicpneumothorax, and occluded vascular grafts) were

missing, and non-registration of such serious complica-tions hampers its database value. The inaccuracy couldbe explained by a great diversity of documented events(n 50), due to a broad definition, suggesting lack ofknowledge of what events should be registered. Inthis aspect, efforts should be put in deciding which com-plications to register, taking severity into account, anddevelop clear definitions for these events (e.g., perfor-mance indicators). Also, training of physicians respon-sible for recording and surveillance methods shouldbe given higher priority. In our institute, a large groupof 12 surgical trainees and nine consultant surgeons

perform identification and documentation of adverseevents, which is probably one cause of missing compli-cations. Although the culture is blame-free and new col-leagues are trained in the registry process, the highnumber of persons responsible for data collection is un-reliable. In this respect, the 37 (19%) of 192 events docu-mented in the registry and not found by the reviewerrepresent an area of inquiry. Analyzing this group ofevents, we concluded that 20 did not meet the criteriain our definition of a complication. This is interestingas it illustrates the lack of understanding by the regis-tering physician. Those missing events appeared to be

negative outcomes (sequel/failure to cure), but mustbe distinguished from complications according to ourdefinition, as has been described by Clavien et al. [18].

As all recorded events are presented at the daily surgi-cal conference/team session, a more careful reviewseems to be necessary to minimize inaccuracies. Fi-nally, a data manager should be introduced for actuallyrecording and coding of the complications. Eight (20%)missing wound infections were discovered after dis-charge. Since there is a general trend towards reducinghospital stay, adequately detecting wound infectionspost-discharge will be a major challenge [19, 20, 21].

Our data suggest that there is a group in which weshould be more alert: patients with longer hospitalstay, acute, complex procedures, and diagnosed in thecategory traumatology are at higher risk for not docu-menting complications in the registry. For instance,trauma patients are at risk for missing a diagnosis dur-ing primary care and, consequently, this could also re-

sult in missing complications [22]. In patients withdiverse pathology, as for instance a polytrauma patient,maintaining overview of the care process is difficult andmost likely responsible for inaccuracies in recordingcomplications. In The Netherlands, national surgicaladverse event registration software has been devel-oped, however it was only installed in 35% of allsurgical departments. This system does not use risk-adjustments and there are concerns about data qualityand public availability [23]. Our registry, as a singlecenter study, has a role in outcome assessment and,for instance in The Netherlands, could be used as

a benchmark. However, in quality of care assessment,risk adjustments and correction for case mix shouldbe made. In this aspect, the National Surgical QualityImprovement Program (ACS_NSQIP) developed bythe American College of Surgeons seems more appro-priate for outcome measurement. This is a validatedprogram that was created to measure and enhancethe care of surgical patients, and has already shownthat the traditional morbidity and mortality conferencesubstantially underreported both in-hospital and post-discharge complications compared with data collectedusing NSQIP techniques [24]. The program is outcome

based, meaning it measures the actual results of opera-tions and has shown that it works as a catalyst for qual-ity improvement [25]. The advantages of this programare that well-defined data are recorded, done by speci-fied trained nurses, resulting in a risk-adjusted predic-tion model for 30 d postoperative mortality andmorbidity. Hannaueret al. demonstrated the feasibilityof using automated processes to improve and extend thecurrent manual review by surgical clinic nurse re-viewers for identifying postoperative complications forthe ACS-NSQIP [26]. Scepticism remains, however,surrounding measurement of individual outcomes and

public reporting [27]. According to the results of thisprogram, adjustment in the areas of definitions, per-sons responsible for recording, and a regular survey ofthe registry are yet discussed by our surgical team.

CONCLUSION

The accuracy of documented complications of the reg-istry appeared to be fairly acceptable, with a substantialinter-rater variability, compared with the rangesmentioned in literature. However, the registry failed

JOURNAL OF SURGICAL RESEARCH: VOL. 173, NO. 1, MARCH 201258


6/6

in adequately documenting some important clinicalcomplications, showed great diversity in type of compli-cations registered (n 50), and 10% of the events re-corded did not meet the criteria of our definition. Ourresults show that if one intends to use a complicationregistry for quality assurance and outcome assessment,development and validating measurement methods

should be made a priority.

REFERENCES

1. Johnston G, Crombie IK, HTODavies HTO, et al. Reviewing au-dit: Barriers and facilitating factors for effective clinical audit.Qual Health Care 2000;9:23.

2. Williams O. What is clinical audit? Ann R Coll Surg Engl 1996;78:406.

3. Bruce J, Russell EM, Mollison J, et al. Measuring and monitor-ing surgical adverse events Health Technol Assess 2001;5:22.

4. Martin RCG II, Brennan MF, Jaques DP. Quality of complica-tion reporting in thesurgical literature. AnnSurg 2002;235:803.

5. Russell EM, Bruce J, Krukowski ZH. Systematic review of the

quality of surgical mortality monitoring. Br J of Surg 2003;90:527.

6. Blank N. High-quality databases: Breaking down barriers. Lan-cet 1999;353:1205.

7. van der MeulenJHP,JacobM, CopleyL. Assessing the quality ofthe data in transplant registry: The European liver transplantregistry. Transplantation 2003;75:2164.

8. Walshe K. Adverse events in health care: Issues in measure-ment. Qual Health Care 2000;9:47.

9. Saklad M. Grading of patients for surgical procedures. Anesthe-siology 1941;2:281.

10. Available at: http://nvvh.artsennet.nl/uri/uriAMGATE_6059_635_TICH_R1647571365637757

11. Veen EJ, Janssen-Heijnen MLG, Leenen LPH, et al. The regis-tration of complications in surgery: A learning curve. World J

Surg 2005;29:402.12. Veen EJ, Steenbruggen J, Roukema JA. Classifying surgicalcomplications: A critical appraisal. Arch Surg 2005;140:1078.

13. Brennan TA, Leape LL, Laird NM, et al. Incidence of adverseevents and negligence in hospitalized patients. Results of theHarvard Medical Practice Study 1. N Eng J Med 1991;324:370.

14. Fine LG, Keogh BE, CretinS, etal. How toevaluateand improvethe quality and credibility of an outcomes database: Validationand feedback study on the UK Cardiac Surgery Experience.BMJ 2003;326:25.

15. Bertges DJ, Shackford SR, Cloud AK, et al. Toward optimal re-cording of surgical complications: Concurrent tracking com-pared to the discharge data set. Surgery 2007;141:19.

16. Gunnarsson U, SeligsohnE, Jestin P, et al. Registration andval-idity of surgical complications in colorectal cancer surgery. Br JSurg 2003;90:454.

17. Dreisler E, Schou L, Adamsen S. Completeness and accuracy ofvoluntary reporting to a national case registry of laparoscopiccholecystectomy. Int J Qual Health Care 2001;13:51.

18. Clavien PA, Sanabria JR, Strasberg SM. Proposed classificationof complications of surgery with examples of utility in cholecys-tectomy. Surgery 1992;11:518.

19. Bruce J, Russell EM, Mollison J, et al. The quality of measure-ment of surgical wound infection as the basis for monitoring:

A systematic review. J Hospital Infect 2001;49:99.20. Smith RL, Bohl JK, McElearney ST, et al. Wound infection after

elective colorectal resection. Ann Surg 2004;239:599.21. Petherick ES, Dalton JE, Moore PJ, et al. Methods for identify-

ing surgical wound infection after discharge from hospital:A surgical review. BMC Infect Dis 2006;6:170.

22. Vles WJ, Veen EJ, Roukema JA, et al. Consequences of delayeddiagnoses in trauma patients: A prospective study. J Am CollSurg 2003;197:596.

23. Marang-van de Mheen PJ, Stadlander MC, Kievit J. Adverseoutcomes in surgical patients: Implementation of a nationwidereporting system. Qual Saf Health Care 2006;15:320.

24. Hutter MM, Rowell KS, Devaney LA, et al. Identification of sur-gical complications and deaths: An assessment of the traditionalsurgical morbidity and mortality conference compared with the

American College of Surgeons National Surgical Quality Im-provement Program. J A Coll Surg 2006;203:618.

25. Katherine SR, Florence ET, Hutter MM, et al. Use of NationalSurgical Quality Improvement Program data as catalyst forquality improvement. Surgery 2007;204:1293.

26. Hannauer DA, Englesbe MJ,Cowan JA Jr,et al. Informatics andthe American College of Surgeons National Quality Improve-ment Program: Automated processes could replace manual re-cord review. J Am Coll Surg 2009;208:37.

27. Neuman HB, Michelassi F, Turner JW, et al. Surroundedby quality metrics: What do surgeons think of ACS-NSQIP? Sur-gery 2009;145:27.

VEEN ET AL.: ACCURACY OF DOCUMENTED COMPLICATIONS IN SURGERY 59
http://nvvh.artsennet.nl/uri/uri%3DAMGATE_6059_635_TICH_R1647571365637757http://nvvh.artsennet.nl/uri/uri%3DAMGATE_6059_635_TICH_R1647571365637757http://nvvh.artsennet.nl/uri/uri%3DAMGATE_6059_635_TICH_R1647571365637757http://nvvh.artsennet.nl/uri/uri%3DAMGATE_6059_635_TICH_R1647571365637757http://nvvh.artsennet.nl/uri/uri%3DAMGATE_6059_635_TICH_R1647571365637757http://nvvh.artsennet.nl/uri/uri%3DAMGATE_6059_635_TICH_R1647571365637757

Registro de Complicaciones

Documents

Transcript of Registro de Complicaciones