Validación y aplicabilidad clínica del comprehensive complication index en una población de pacientes intervenidos en un servicio de cirugía general y del aparato digestivo. Estudio prospectivo
- Autores: Roberto de la Plaza Llamas
- Directores de la Tesis: Juan Manuel Bellón Caneiro (dir. tes.), José Manuel Ramia Ángel (codir. tes.)
- Lectura: En la Universidad de Alcalá ( España ) en 2018
- Idioma: español
- Tribunal Calificador de la Tesis: Javier Arias Díaz (presid.), Natalio García Honduvilla (secret.), Damián García Olmo (voc.)
- Programa de doctorado: Programa de Doctorado en Ciencias de la Salud por la Universidad de Alcalá
- Materias:
- Enlaces
- Resumen
- español
Las complicaciones postoperatorias suponen un parámetro clave para medir los resultados de los procedimientos quirúrgicos. Sin embargo, su cuantificación no se lleva a cabo habitualmente en los servicios quirúrgicos. En el mejor de los casos, se miden en cirugías concretas y durante un tiempo limitado. No se puede determinar qué servicios lo hacen mejor y cuáles son merecedores de ser los benchmarking. Sólo el conocimiento real de la morbilidad nos puede permitir implementar medidas de mejora.
Los objetivos primarios de este trabajo fueron la validación en términos clínicos, de discapacidad, reingreso, estancia hospitalaria postoperatoria y prolongación de la estancia de la Clasificación de Clavien Dindo (CDC), del Comprehensive Complication Index (CCI) y del Complication Severity Score (CSS). Objetivos secundarios fueron entre otros, medir la frecuencia de complicaciones que ocurren en todos pacientes sometidos a intervención quirúrgica en un servicio de cirugía general, las variables que se asocian y el momento en que tienen lugar.
En cuanto a la metodología, se incluyeron prospectivamente todos los pacientes cuya primera intervención quirúrgica se produjo entre el 1 de marzo de 2016 y el 28 de febrero de 2017, en el Servicio de Cirugía General y del Aparato Digestivo del Hospital Universitario de Guadalajara. Se creó un formulario específico en la historia electrónica Mambrino XXI®. Se analizaron todas las complicaciones asociadas a la cirugía del ingreso inicial en los primeros 90 días desde la cirugía o hasta que el enfermo fue dado de alta de cualquier ingreso producido con anterioridad a los 90 días. Para ello se consultó el formulario descrito, las notas de evolución médicas y de enfermería. Las complicaciones fueron clasificadas según la CDC y se calculó el CCI y el CSS. Para homogeneizar la muestra, se clasificaron las intervenciones quirúrgicas según el Operative Severity Score (OSS) descrito por Copeland et al. en 1991: menor, moderada, mayor y mayor+. Finalmente se incluyeron en el estudio 1850 pacientes. Se produjo complicaciones en 513 pacientes (27.7%), de ellos 249 (48.5%) presentaron 1 complicación y 264 (51.5%) más de 1. Fallecieron 42 (2.3%) pacientes. Según el OSS: 777 (42%) operaciones fueron menores, 625 (33.8%) moderadas, 392 mayores (21.2%) y 56 (3%) mayores+. Hubo un incremento adicional del 11.6% si el computo de la morbilidad se hizo a los 90 días frente a los 30 días. Reingresaron 101 (5.5%) de los pacientes.
La CDC, el CCI y el CSS se asociaron significativamente con todos los modelos de regresión logística (discapacidad, reingreso, prolongación de la cirugía), regresión lineal (estancia hospitalaria postoperatoria) y grupos de complejidad de la cirugía. Además, los tres índices estuvieron muy correlacionados ( > 94%).
De forma comparativa, en general, la CDC mostró unos valores de validación clínica discretamente inferiores a los índices CCI y CSS. Las diferencias entre ambos fueron mínimas.
En nuestra serie el CCI y CSS aportaron información adicional respecto al CDC en el 51% de los pacientes que presentaron complicaciones.
EL CCI o el CSS son herramientas necesarias en la valoración y comparación de la morbilidad postoperatoria en los servicios de Cirugía General.
- English
Postoperative complications represent a key parameter for measuring the results of surgical procedures, and as such constitute an essential indicator of quality of care. Nonetheless, they are not regularly quantified at surgery services; and in the cases in which they are measured, the assessment tends to focus on specific surgical techniques and on limited time periods. One might think that this quantification should be prospective, mandatory and global, especially since its consequences directly influence patients’ quality of life, prognosis, quality of care and economic costs. In the absence of an assessment of this kind it is difficult to establish the real morbidity associated with a specific procedure at a specific service, and so it is hard to determine which services perform best and can serve as benchmarks for the others. We need an accurate picture of postoperative morbidity in order to be able to implement improvements in surgical care.
In addition to the reluctance to record and report complications, until recently no standardized classification system has been available for their registration and comparison. In 2004, Dindo et al. published a system which later became known as the Clavien-Dindo Classification (CDC) and has been widely used in many fields of surgery. Currently, the 2004 article has 8167 citations in the Web of Science. Although the CDC is a good system for grading complications, a major disadvantage is the fact that the entire postoperative course is defined only by the most serious complication, with others of lesser magnitude being ignored. In response to this problem, in 2013 Slankamenac et al. developed a new global scoring system for postoperative complications based on the CDC, which they called the Comprehensive Complication Index (CCI). In the CCI, all complications are evaluated separately according to the CDC and are then entered in the online calculator ttp://www.assessurgery.com/about_cci-calculator/ to quantify the morbidity on a scale from 0 to 100 (a score of 0 reflects zero complications, while one of 100 indicates that the patient has died). This paper currently has 73 references in PubMed. Although the CCI has been used for specific operations, to date it has not been applied to the entire range of pathologies presented at a general surgery service. In 2015, the Complication Severity Score (CSS) was published, also based on the CDC, and again with a global score of 0 to 100. Its authors contend that this score performs better than the CCI, but that it needs to be validated clinically.
This PhD thesis poses the following question: Does the Comprehensive Complication Index offer a better summary of the postoperative complications and their clinical outcomes in patients operated at a general and digestive surgery service than the Clavien Dindo Classification or the Complication Severity Score? The primary objectives were the clinical validation of disability, re-admission, postoperative hospital stay and prolongation of hospital stay of each of the indices analysed (CDC, CCI and CSS). Secondary objectives were to measure the frequency of complications in all patients undergoing surgery at a general surgery service, the associated variables and when they occur.
As regards the study’s methodology, all patients operated upon for the first time between March 1, 2016 and February 28, 2017 at the General Surgery and Digestive Diseases Service of the University Hospital of Guadalajara were included. Patients undergoing minor outpatient surgery were excluded. All the complications presented by the patients undergoing surgery were recorded in the clinical history. Physicians were also trained in the use of the CDC and in the calculation of the CCI, and a form was created using the Mambrino XXI® electronic medical record. By consulting this form and the medical and nursing evolution notes, we analysed all the complications associated with the initial surgery occurring within 90 days or, in the cases of re-admission within 90 days, until the patient was discharged. Complications were classified according to the CDC, and the CCI and the CSS were calculated. To standardize the sample, surgical interventions were classified as minor, moderate, major and major + according to the Operative Severity Score (OSS, Copeland et al. 1991). Hospital stay was considered prolonged if it was above 75% of the stay of patients with the longest admission time for each of the OSS groups.
Finally, 1850 patients were included in the study. There were complications in 513 patients (27.7%), and 42 (2.3%) patients died. According to the OSS: 777 (42%) operations were minor, 625 (33.8%) moderate, 392 major (21.2%) and 56 (3%) major +. There was an additional increase of 11.6% if the morbidity was computed at 90 days rather than 30 days. One hundred and one patients (5.5%) were readmitted.
The dependent variable “presence of some complication” was significantly associated in the univariate analysis (p <0.001) with the following independent variables: prolonged hospital stay (in the minor and moderate surgery groups) and days of admission. It was also associated with age, ASA, type of anaesthesia, surgery time, surgery shift, approach, surgery complexity and surgeon. It was not associated with age or surgery scheduling (i.e., elective or emergency).
At the end of the follow-up, 249 (48.5%) of the 513 patients with complications presented one complication, and 264 (51.5%) more than one. The presence of two or more complications was more frequent after more complex surgeries and also with increased CDC grade (with the exception of grade V).
With regard to the validation of the indices, we observed the following:
• The CDC, the CCI and the CSS were significantly associated with all the logistic regression models (disability, readmission, prolongation of the surgery), linear regression (postoperative hospital stay) and complexity of surgery.
• The CDC was influenced by a greater number of confounding factors than the CCI and the CSS in the different models.
• Discrimination (AUC-ROC) was high in the disability and re-admission models and prolongation of the stay in major and major + surgery. It was fair in minor surgery (<0.7) in all three indices and moderate in the CCI.
• Calibration was poor in the disability and re-admission models.
• The percentage of variability that explains the adjusted R2 ranged between 33.90% and 69.40% for the four surgical complexity groups. The CDC had slightly worse results in terms of the adjusted R2 than the CCI: the CSS improved slightly on the CCI, except in the moderate surgery group.
The CDC, CCI and CSS were highly correlated (> 94%).
As conclusions, we can affirm that:
1. It cannot be categorically stated that the CCI provides a better summary of the postoperative complications and their clinical results than the CDC and the CSS.
2. The CDC, the CCI and the CSS presented significant associations with all the logistic regression models (disability, readmission, prolongation of the surgery), linear regression (postoperative hospital stay) and complexity of surgery.
3. In general, the CDC’s clinical validation values were slightly lower than those of the CCI and CSS. The differences between the CCI and the CSS were minimal.
4. In our series, the CCI and the CSS provided additional information not supplied by the CDC in 51% of patients who presented complications.
5. Assessing complications at 90 days post-surgery instead of at 30 days increased their identification by 11%.
6. The CCI and the CSS are necessary tools in the assessment and comparison of postoperative morbidity at general surgery services.
- español
