Pancreatoyeyunostomía conducto‐mucosa versus otros tipos de pancreatoyeyunostomía para la prevención de la fístula pancreática posoperatoria después de la pancreatoduodenectomía
Resumen
Antecedentes
La fístula pancreática posoperatoria es una complicación frecuente y grave después de la pancreatoduodenectomía. La pancreatoyeyunostomía conducto‐mucosa se ha utilizado en muchos centros para reconstruir la continuidad digestiva pancreática tras una pancreatoduodenectomía. Sin embargo, su eficacia y seguridad no están claras.
Objetivos
Evaluar los efectos beneficiosos y perjudiciales de la pancreatoyeyunostomía conducto‐mucosa versus otros tipos de pancreatoyeyunostomía para la reconstrucción de la continuidad digestiva pancreática en participantes sometidos a pancreatoduodenectomía, y comparar los efectos de las diferentes técnicas de pancreatoyeyunostomía conducto‐mucosa.
Métodos de búsqueda
Se realizaron búsquedas en la Biblioteca Cochrane (2021, número 1), MEDLINE (1966 hasta el 9 de enero de 2021), Embase (1988 hasta el 9 de enero de 2021) y Science Citation Index Expanded (1982 hasta el 9 de enero de 2021).
Criterios de selección
Se incluyeron todos los ensayos controlados aleatorizados (ECA) que compararon la pancreatoyeyunostomía conducto‐mucosa con otros tipos de pancreatoyeyunostomía (p. ej., pancreatoyeyunostomía de invaginación, pancreatoyeyunostomía por unión) en participantes sometidos a pancreatoyeyunectomía. También se incluyeron los ECA que compararan diferentes tipos de pancreatoyeyunostomía conducto‐mucosa en participantes sometidos a pancreatoduodenectomía.
Obtención y análisis de los datos
Dos autores de la revisión, de forma independiente, identificaron los estudios para la inclusión, recopilaron los datos y evaluaron el riesgo de sesgo. El metanálisis se realizó con el programa informático Review Manager 5. Se calculó la razón de riesgos (RR) para los desenlaces dicotómicos y la diferencia de medias (DM) para los desenlaces continuos con intervalos de confianza (IC) del 95%. Para todos los análisis, se utilizó el modelo de efectos aleatorios. Se utilizó la herramienta Cochrane RoB 1 para evaluar el riesgo de sesgo. Se utilizó el método GRADE para evaluar la certeza de la evidencia de todos los desenlaces.
Resultados principales
En la revisión se incluyeron 11 ECA con un total de 1696 participantes. Un ECA fue un estudio que incluyó dos centros; los otros diez ECA incluyeron un solo centro y se realizaron en: China (cuatro estudios); Japón (dos estudios); EE.UU. (un estudio); Egipto (un estudio); Alemania (un estudio); India (un estudio); e Italia (un estudio). La media de edad de los participantes varió entre 54 y 68 años. Todos los ECA tuvieron un alto riesgo de sesgo.
Pancreatoyeyunostomía conducto‐mucosa versus cualquier otro tipo de pancreatoyeyunostomía
Se incluyeron diez ECA con 1472 participantes que compararon la pancreatoyeyunostomía conducto‐mucosa con la pancreatoyeyunostomía de invaginación: 732 participantes fueron asignados al azar al grupo de conducto‐mucosa y 740 participantes fueron asignados al azar al grupo de invaginación después de la pancreatoduodenectomía. Al comparar las dos técnicas, no está muy clara la evidencia acerca de la tasa de fístula pancreática posoperatoria (grado B o C; RR 1,45; IC del 95%: 0,64 a 3,26; siete estudios, 1122 participantes; evidencia de certeza muy baja), la mortalidad posoperatoria (RR 0,77; IC del 95%: 0,39 a 1,49; diez estudios, 1472 participantes; evidencia de certeza muy baja), la tasa de reintervención quirúrgica (RR 1,12; IC del 95%: 0,65 a 1,95; diez estudios, 1472 participantes; evidencia de certeza muy baja), la tasa de hemorragia posoperatoria (RR 0,85; IC del 95%: 0,51 a 1,42; nueve estudios, 1275 participantes; evidencia de certeza muy baja), la tasa general de complicaciones quirúrgicas (RR 1,12; IC del 95%: 0,92 a 1,36; cinco estudios, 750 participantes; evidencia de certeza muy baja) ni la duración de la estancia hospitalaria (DM ‐0,41 días; IC del 95%: ‐1,87 a 1,04; cuatro estudios, 658 participantes; evidencia de certeza muy baja). Los estudios no informaron acerca de los eventos adversos ni de los desenlaces de la calidad de vida.
Un tipo de pancreatoyeyunostomía conducto‐mucosa versus otro tipo de pancreatoyeyunostomía conducto‐mucosa
Se incluyó un ECA con 224 participantes que comparó la pancreatoyeyunostomía conducto‐mucosa mediante la técnica de Blumgart modificada con la pancreatoyeyunostomía conducto‐mucosa mediante la técnica interrumpida tradicional: 112 participantes fueron asignados al azar al grupo de Blumgart modificado, y 112 participantes fueron asignados al azar al grupo de técnica interrumpida tradicional después de la pancreatoduodenectomía. Al comparar las dos técnicas, no está muy clara la evidencia en cuanto a la tasa de fístula pancreática posoperatoria (grado B o C; RR 1,51; IC del 95%: 0,61 a 3,75; un estudio, 210 participantes; evidencia de certeza muy baja), la mortalidad posoperatoria (no hubo muertes en ninguno de los dos grupos; un estudio, 210 participantes; evidencia de certeza muy baja), la tasa de reintervención quirúrgica (RR 1,93; IC del 95%: 0,18 a 20,91; un estudio, 210 participantes; evidencia de certeza muy baja), tasa de hemorragia posoperatoria (RR 2,89; IC del 95%: 0,12 a 70,11; un estudio, 210 participantes; evidencia de certeza muy baja), la tasa general de complicaciones quirúrgicas (RR 1,10; IC del 95%: 0,80 a 1,51; un estudio, 210 participantes; evidencia de certeza muy baja) ni la duración de la estancia hospitalaria (15 días versus 15 días; un estudio, 210 participantes; evidencia de certeza muy baja). El estudio no informó acerca de los eventos adversos ni de los desenlaces de la calidad de vida.
Conclusiones de los autores
No está muy clara la evidencia acerca de la los efectos de la pancreatoyeyunostomía conducto‐mucosa en comparación con la pancreatoyeyunostomía de invaginación sobre cualquiera de los desenlaces, incluida la tasa de fístula pancreática posoperatoria (grado B o C), la mortalidad posoperatoria, la tasa de reintervención quirúrgica, la tasa de hemorragia posoperatoria, la tasa general de complicaciones quirúrgicas y la duración de la estancia hospitalaria. Tampoco está muy clara la evidencia en cuanto a si la pancreatoyeyunostomía conducto‐mucosa mediante la técnica de Blumgart modificada es superior, equivalente o inferior a la pancreatoyeyunostomía conducto‐mucosa mediante la técnica interrumpida tradicional. Ninguno de los estudios informó sobre los eventos adversos ni los desenlaces de calidad de vida.
PICO
Resumen en términos sencillos
¿La fijación del conducto excretor del páncreas a la segunda parte del intestino delgado es mejor que otros métodos de reconstrucción para reducir la fuga de jugo del páncreas a los tejidos abdominales?
Mensajes clave
‐ No se sabe si la pancreatoyeyunostomía conducto‐mucosa (unión del conducto excretor del páncreas a la segunda parte del intestino delgado) reduce la fístula pancreática posoperatoria (fuga de jugo del páncreas a los tejidos abdominales) en comparación con la pancreatoyeyunostomía de invaginación (inserción del páncreas dentro de la segunda parte del intestino delgado).
‐ No se sabe si la pancreatoyeyunostomía conducto‐mucosa modificada es superior, equivalente o inferior a la pancreatoyeyunostomía conducto‐mucosa tradicional.
‐ Los estudios futuros deberían utilizar métodos adecuados para mejorar la confianza en la evidencia.
¿Qué es una fístula pancreática posoperatoria?
El páncreas es una glándula digestiva situada en la parte posterior del abdomen superior. También es vital para el control normal del azúcar en la sangre. El tratamiento quirúrgico estándar para el cáncer o la inflamación del páncreas es la extirpación parcial de la cabeza del páncreas, junto con el intestino cercano, mediante un procedimiento conocido como pancreatoduodenectomía. La pancreatoduodenectomía consiste en reconectar el páncreas y la segunda parte del intestino delgado (un procedimiento conocido como pancreatoyeyunostomía) para permitir que el jugo pancreático que contiene enzimas digestivas entre en el sistema digestivo. Una fístula pancreática posoperatoria se produce cuando la reconexión no cicatriza correctamente, creando una fuga de jugo pancreático del páncreas a los tejidos abdominales. La fístula pancreática posoperatoria es una complicación que retrasa la recuperación de la cirugía y a menudo requiere un tratamiento adicional para garantizar la curación completa.
¿Qué se puede hacer para reducir la fístula pancreática posoperatoria?
Los métodos de reconexión del páncreas y la segunda parte del intestino delgado en las personas sometidas a pancreatoduodenectomía incluyen:
‐pancreatoyeyunostomía conducto‐mucosa;
‐pancreatoyeyunostomía de invaginación;
‐pancreatoyeyunostomía por unión (unión del páncreas y la segunda parte del intestino delgado).
La pancreatoyeyunostomía conducto‐mucosa es un método que se utiliza habitualmente en todo el mundo para reducir la fístula pancreática posoperatoria después de la pancreatoduodenectomía. Sin embargo, la seguridad y la efectividad de la pancreatoyeyunostomía conducto‐mucosa todavía no está clara.
¿Qué se quería averiguar?
Se quería averiguar si la pancreatoyeyunostomía conducto‐mucosa es mejor que cualquier otro tipo de pancreatoyeyunostomía para mejorar:
‐ la tasa de fístula pancreática posoperatoria (proporción de participantes con fístula pancreática posoperatoria);
‐ la tasa de mortalidad.
También se quería averiguar si la pancreatoyeyunostomía conducto‐mucosa se asocia con algún efecto no deseado.
¿Qué se hizo?
Primero se buscaron estudios que compararan:
‐ la pancreatoyeyunostomía conducto‐mucosa frente a cualquier otro tipo de pancreatoyeyunostomía; o
‐ diferentes tipos de pancreatoyeyunostomía conducto‐mucosa tras una pancreatoduodenectomía.
Se compararon y resumieron los resultados de los estudios y la confianza en la evidencia se evaluó sobre la base de factores como la metodología y el tamaño de los estudios.
¿Qué se encontró?
Se encontraron diez estudios con 1472 personas en los que se comparó la pancreatoyeyunostomía conducto‐mucosa con la pancreatoyeyunostomía de invaginación en personas sometidas a pancreatoduodenectomía. El estudio más grande incluyó 308 personas, y el más pequeño 64. Los estudios se realizaron en países de todo el mundo; la mayoría se hicieron en China (cuatro). La mayoría de los estudios duró aproximadamente dos años; solo tres estudios duraron cuatro años o más. Cuatro de los estudios estuvieron financiados por subvenciones no comerciales. Los resultados de estos estudios no permiten saber si la pancreatoyeyunostomía conducto‐mucosa reduce:
‐ tasa de fístula pancreática posoperatoria, o
‐ la tasa de mortalidad.
Los estudios no informaron acerca de los efectos no deseados.
Solo se encontró un estudio de un único centro con 224 personas en el que la pancreatoyeyunostomía conducto‐mucosa modificada se comparó con la pancreatoyeyunostomía conducto‐mucosa tradicional en personas sometidas a pancreatoduodenectomía. El estudio se realizó en Japón y duró aproximadamente cuatro años. El estudio no informó sus fuentes de financiación. A partir de los resultados de este estudio no se puede señalar si la pancreatoyeyunostomía conducto‐mucosa modificada reduce:
‐ tasa de fístula pancreática posoperatoria, o
‐ la tasa de mortalidad.
El estudio no informó efectos no deseados.
¿Cuáles son las limitaciones de la evidencia?
Se tiene muy poca confianza en la evidencia porque la mayoría de los estudios incluidos presentaban algunas limitaciones en cuanto a la forma en que se realizaron o informaron.
¿Cuál es el grado de actualización de esta evidencia?
La evidencia está actualizada hasta enero de 2021.
Authors' conclusions
Summary of findings
| Duct‐to‐mucosa compared to invagination pancreaticojejunostomy for the prevention of postoperative pancreatic fistula following pancreaticoduodenectomy | ||||||
| Patient or population: adults undergoing open pancreaticoduodenectomy | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with invagination pancreaticojejunostomy | Risk with duct‐to‐mucosa pancreaticojejunostomy | |||||
| Rate of POPF (grade B or C defined according to the 2005 ISGPF definition) Follow‐up: 30 days | 80 per 1000 | 115 per 1000 | RR 1.45 | 1122 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of POPF (grade B or C defined according to the 2005 ISGPF definition). |
| Postoperative mortality Follow‐up: 30 days | 26 per 1000 | 20 per 1000 | RR 0.77 | 1472 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on postoperative mortality. |
| Adverse events | This outcome was not reported in any of the included studies. | |||||
| Rate of surgical reintervention Follow‐up: 30 days | 34 per 1000 | 38 per 1000 | RR 1.12 | 1472 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of surgical reintervention. |
| Rate of postoperative bleeding Follow‐up: 30 days | 47 per 1000 | 40 per 1000 | RR 0.85 | 1275 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of postoperative bleeding. |
| Overall rate of surgical complications Follow‐up: 30 days | 484 per 1000 | 542 per 1000 | RR 1.12 | 750 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the overall rate of surgical complications. |
| Length of hospital stay (days) | The mean length of hospital stay was 16 to 22 days | The mean length of hospital stay was 14.9 to 20 days | MD ‐0.41 (‐1.87 to 1.04) | 658 | ⊕⊝⊝⊝ | Six additional studies with skewed data were not suitable for pooling. One trial reported that the duct‐to‐mucosa group was associated with less time in hospital, but another trial reported that the duct‐to‐mucosa group was associated with more time in hospital than the invagination group. The other four studies reported no difference in the length of hospital stay between groups. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the length of hospital stay. |
| Quality of life | This outcome was not reported in any of the included studies. | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded one level for serious risk of bias: all included studies with high risk of bias for blinding of participants and personnel and some of the studies had high risk of bias for blinding of outcome assessment. | ||||||
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy usingthe traditional interrupted technique for the prevention of postoperative pancreatic fistula following pancreaticoduodenectomy | ||||||
| Patient or population: adults undergoing open pancreaticoduodenectomy | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Risk with duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique | |||||
| Rate of POPF (grade B or C defined according to the 2016 ISGPS definition) Follow‐up: 30 days | 68 per 1000 | 103 per 1000 | RR1.51 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of POPF (grade B or C defined according to the 2016 ISGPS definition) compared with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique. |
| Postoperative mortality Follow‐up: 30 days | See comment | See comment | Not estimable | 210 | ⊕⊝⊝⊝ | There were no postoperative deaths in either group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on postoperative mortality. |
| Adverse events | This outcome was not reported in the study. | |||||
| Rate of surgical reintervention Follow‐up: 30 days | 10 per 1000 | 19 per 1000 | RR 1.93 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of surgical reintervention. |
| Rate of postoperative bleeding Follow‐up: 30 days | ‐ | ‐ | RR 2.89 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of postoperative bleeding. |
| Overall rate of surgical complications Follow‐up: 30 days | 408 per 1000 | 449 per 1000 | RR 1.10 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the overall rate of surgical complications . |
| Length of hospital stay (days) | The median length of hospital stay was 15 days | The median length of hospital stay was 15 days | See comment | 210 | ⊕⊝⊝⊝ | Authors reported that there was no difference in the length of hospital stay between groups. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the length of hospital stay. |
| Quality of life | This outcome was not reported in the study. | |||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded two levels for very serious risk of bias: the study was at high risk of bias for blinding of participants and personnel, blinding of outcome assessment, and incomplete outcome data. | ||||||
Background
See Appendix 1 for a glossary of terms.
Description of the condition
Pancreatic cancer is the fourteenth most common cancer and the eighth leading cause of cancer mortality worldwide (Bray 2018; Ferlay 2019; Mizrahi 2020; Park 2021; Siegel 2021). The numbers of new cases and deaths from pancreatic cancer are increasing from a global viewpoint, accounting for about 448,000 new cases and 441,000 deaths in 2017 (GBD 2019). The global incidence of pancreatic cancer was 5 cases per 100,000 people per year in 1990, increasing to approximately 6 cases per 100,000 people per year in 2017 (Bray 2018; GBD 2019). Although the exact causes of pancreatic cancer are still unclear, the common risk factors for development of pancreatic cancer are obesity, diabetes, and heavy tobacco usage (GBD 2019; Maisonneuve 2019; Mizrahi 2020).
Pancreatoduodenectomy is a major surgical operation used to treat pancreatic, biliary, ampullary, and duodenal cancers, and other diseases, such as chronic pancreatitis (Cheng 2014; Cheng 2017; Deng 2020; Gurusamy 2013; Hüttner 2016). Pancreatoduodenectomy carries a 30% to 60% overall surgical complication rate, but a low postoperative mortality rate (less than 5%) (Bassi 2005; Chen 2015; Chou 1996; Connor 2005; Gurusamy 2013). The most common and potentially life‐threatening complication of pancreatoduodenectomy is postoperative pancreatic fistula (POPF; Cheng 2017; Deng 2020; Dong 2016; Gurusamy 2013).
The International Study Group on Pancreatic Fistula (ISGPF) proposed a definition of POPF by consensus in 2005 (Bassi 2005). POPF is defined by the ISGPF as "a drain output of any measurable volume of fluid on or after postoperative day 3 with an amylase content greater than 3 times the serum amylase activity" (amylase is an important enzyme that is present in pancreatic juice, and a high level of amylase in abdominal collections indicates POPF) (Bassi 2005). POPF has been graded as A, B, and C (Bassi 2005). According to the 2005 ISGPF definition, grade A POPF is a transient fistula, with no clinical impact (Bassi 2005). Grade B POPF requires a change in the management of the patient. Grade C POPF requires a major change in the management of the patient and usually requires reoperation (Bassi 2005). There are some unclear distinctions between grade B POPF and grade C POPF in the 2005 ISGPF definition. For example, the placement of an invasive drain with a percutaneous approach was defined as grade C in the text, whereas it was considered a “possible” grade B (“yes/no”) procedure in the summary table (Bassi 2005). As a result, patients undergoing percutaneous drainage were often classified differently, according to the various interpretations.
The International Study Group on Pancreatic Surgery (ISGPS) updated the definition of POPF in 2016 to be "a drain output of any measurable volume of fluid with an amylase level > 3 times the upper limit of institutional normal serum amylase activity, associated with a clinically relevant development/condition related directly to the POPF" (Bassi 2017). The definition of grade B and C POPF has also been clarified in the 2016 ISGPS definition: "Grade B POPF requires a change in the postoperative management; drains are either left in place > 3 weeks or repositioned through endoscopic or percutaneous procedures. Grade C POPF requires reoperation or leads to single or multiple organ failure and/or mortality attributable to the POPF" (Bassi 2017). The 2016 ISGPS definition provided a specific definition of organ failure as “the need for reintubation, hemodialysis, and/or use of inotropic agents for > 24 hours because of respiratory, renal, or cardiac insufficiency, respectively”; sepsis was excluded from the 2016 ISGPS definition of organ failure (Bassi 2017).
POPF often occurs from the dehiscence (leakage) of the pancreatic‐enteric anastomosis following pancreaticoduodenectomy (Bassi 2005; Hackert 2011). The rate of Grade B or C POPF is as high as 10% to 20% (McMillan 2016; Rivas 2019). Many risk factors for POPF have been identified, including pancreatic pathology, pancreatic texture, pancreatic duct size, and intraoperative blood loss (Callery 2013; McMillan 2016). People with ampullary, duodenal, cystic, or islet cell diseases, soft pancreatic textures, small pancreatic duct sizes (e.g. < 3 mm), or massive intraoperative blood losses (e.g. > 1000 mL) seem to be more likely than others to experience POPF (Callery 2013; McMillan 2016).
Description of the intervention
Many methods have been introduced to reduce the POPF rate following pancreaticoduodenectomy, such as insertion of pancreatic duct stents (Dong 2016), administration of somatostatin or its analogues (Gurusamy 2013), application of fibrin sealants (Deng 2020), Roux‐en‐Y reconstruction with isolated pancreaticojejunostomy (Klaiber 2015), and pancreaticogastrostomy instead of pancreaticojejunostomy (Cheng 2017; Xiong 2014). At present, duct‐to‐mucosa pancreaticojejunostomy is the most commonly used method of reconstruction worldwide for the prevention of POPF following pancreaticoduodenectomy (Conzo 2015; Lai 2009; Testini 2016).
When performing a duct‐to‐mucosa pancreaticojejunostomy, a small opening is made in the jejunum that matches the size of the main pancreatic duct (Bai 2016; Senda 2018; Testini 2016). Consequently, the jejunal mucosa is exposed to the main pancreatic duct; then, the main pancreatic duct is anastomosed to the jejunal mucosa in all directions. Finally, the pancreatic stump, which includes other branch pancreatic ducts, is anastomosed to the jejunal serosa (Bai 2016; Senda 2018; Testini 2016).
How the intervention might work
Duct‐to‐mucosa pancreaticojejunostomy, which was first introduced by Varco in 1945 (Varco 1945), has been advocated by many surgeons for the reconstruction of pancreatic digestive continuity following pancreatoduodenectomy (Chou 1996). The primary reasons for performing duct‐to‐mucosa pancreaticojejunostomy rather than using another technique are as follows: 1) direct anastomosis between the jejunal mucosa and the pancreatic duct is beneficial for healing of the pancreatic‐enteric anastomosis; 2) the pancreatic remnant is protected by the jejunal wall; 3) the main pancreatic duct is anastomosed to the jejunal mucosa, and the other branch pancreatic ducts are closed by the jejunal serosa; and 4) duct‐to‐mucosa pancreaticojejunostomy can drain the main duct into the jejunum and preserve the patency of the duct (El Nakeeb 2015; Hua 2015; Kilambi 2018; Zhang 2017). Due to these advantages, duct‐to‐mucosa pancreaticojejunostomy has the potential to prevent POPF.
However, performing duct‐to‐mucosa pancreaticojejunostomy for people with small pancreatic ducts is difficult and time consuming. Thus, alternative techniques have been introduced to reduce the rate of POPF, such as invagination pancreaticojejunostomy and binding pancreaticojejunostomy (Peng 2007; Senda 2018).
Invagination pancreaticojejunostomy, also known as dunking pancreaticojejunostomy, is another reconstruction technique commonly employed during pancreatoduodenectomy (Shrikhande 2017). Invagination pancreaticojejunostomy is relatively easy to learn and to perform as there is no need to identify the main pancreatic duct. Both the pancreatic stump and the pancreatic ducts are invaginated into the jejunum to prevent POPF (Rivas 2019; Xiang 2019). However, long‐term direct exposure of the pancreatic stump to the digestive juice in the jejunum may induce pancreatic stump necrosis and pancreatic duct stenosis (Rivas 2019; Xiang 2019). In addition, forcibly invaginating the pancreatic stump into the jejunum may cause ischemia of the pancreatic stump and jejunum, which increases the risk of POPF (Kone 2020; Rivas 2019; Xiang 2019).
Binding pancreaticojejunostomy, first introduced by Peng in 2002 (Peng 2002), is a novel reconstruction technique that can be used during pancreatoduodenectomy. This technique does not require needle holes on the surface of the pancreas (Peng 2002). In addition, the gap between the jejunal wall and the pancreas stump is eliminated after binding, and this helps prevent POPF. However, it is very difficult to control the binding intensity. Binding that is too loose causes POPF, while binding that is too tight induces ischemia of the pancreatic stump and jejunum, which also results in POPF (Casadei 2013; Xiang 2019). Furthermore, it is difficult to perform binding pancreaticojejunostomy when the jejunum is too small or the pancreatic stump is too large (Peng 2002; Xiang 2019).
Why it is important to do this review
The choice of reconstruction method to be used in people undergoing pancreaticoduodenectomy is controversial. There are currently no international guidelines on how to reconstruct pancreatic digestive continuity after pancreaticoduodenectomy (Shrikhande 2017). Up until now, there has been no Cochrane Review assessing the benefits and harms of duct‐to‐mucosa pancreaticojejunostomy for the reconstruction of pancreatic digestive continuity following pancreaticoduodenectomy.
Objectives
To assess the benefits and harms of duct‐to‐mucosa pancreaticojejunostomy versus other types of pancreaticojejunostomy for the reconstruction of pancreatic digestive continuity in participants undergoing pancreaticoduodenectomy, and to compare the effects of different duct‐to‐mucosa pancreaticojejunostomy techniques.
Methods
Criteria for considering studies for this review
Types of studies
We included all randomized controlled trials (RCTs)— regardless of sample size, language, or publication status — which compared duct‐to‐mucosa pancreaticojejunostomy with other types of pancreaticojejunostomy, or which compared different types of duct‐to‐mucosa pancreaticojejunostomy in people undergoing pancreaticoduodenectomy. We excluded quasi‐randomized studies, in which the allocation was performed on the basis of a pseudorandom sequence (e.g. odd or even hospital number or date of birth, alternation), and non‐randomized studies, because of the potential for bias (Reeves 2021).
Types of participants
We included adults (irrespective of sex or race) who were diagnosed with any pancreatic/duodenal/periampullary disease requiring pancreatoduodenectomy.
Types of interventions
We assessed the following comparisons for people undergoing pancreatoduodenectomy.
-
Duct‐to‐mucosa pancreaticojejunostomy versus any other type of pancreaticojejunostomy (e.g. invagination pancreaticojejunostomy, binding pancreaticojejunostomy)
-
One type of duct‐to‐mucosa pancreaticojejunostomy versus a different type of duct‐to‐mucosa pancreaticojejunostomy (e.g. double layers versus triple layers)
We included any additional interventions, provided they were not part of the randomized treatment.
Types of outcome measures
Primary outcomes
-
Rate of POPF (measured at 30 days; grade B or C). We accepted both the 2005 ISGPF definition and the 2016 ISGPS definition (Bassi 2005; Bassi 2017)
-
Postoperative mortality (measured at 30 days)
-
Adverse events (measured at 90 days; number of people with at least one adverse event within 3 months). We accepted all adverse events reported by the study authors, irrespective of the severity of the adverse events.
Secondary outcomes
-
Rate of surgical reintervention (measured at 30 days; to repair a pancreatic fistula, drain an intra‐abdominal abscess, or stop bleeding)
-
Rate of postoperative bleeding (measured at 30 days)
-
Overall rate of surgical complications (measured at 30 days; classified according to the Clavien‐Dindo classification of surgical complications (Clavien 2009; Dindo 2004))
-
Length of hospital stay
-
Quality of life (measured at 30 days using any validated scale)
The main reason to justify duct‐to‐mucosa pancreaticojejunostomy is the assumption that it will reduce the rate of POPF; therefore we chose the rate of POPF as our primary outcome. We chose other clinical outcomes to assess whether duct‐to‐mucosa pancreaticojejunostomy resulted in lower rates of death, surgical reintervention, postoperative bleeding, and overall surgical complications, earlier discharge from the hospital, or improvement in health‐related quality of life.
Reporting of the outcomes listed here was not an inclusion criterion for the review.
Search methods for identification of studies
We designed the search strategies with the help of the Cochrane Gut Information Specialist. No restrictions were placed on the language of publication when searching the electronic databases, or reviewing reference lists in identified studies.
Electronic searches
We conducted a literature search to identify all published and unpublished RCTs. The literature search did not restrict by language of publication. We translated the non‐English language papers into English and fully assessed them for potential inclusion in the review as necessary.
We searched the following electronic databases for identifying potential studies.
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2021, Issue 1) (Appendix 2)
-
MEDLINE (1966 to 9 January 2021) (Appendix 3)
-
Embase (1988 to 9 January 2021) (Appendix 4)
-
Science Citation Index Expanded (1982 to 9 January 2021) (Appendix 5)
Searching other resources
We checked the reference lists of all primary studies and review articles for additional references. We contacted the authors of the identified studies and asked them to identify other published and unpublished studies. We also contacted experts in the field.
We searched for errata or retractions from eligible studies on PubMed (www.ncbi.nlm.nih.gov/pubmed) and reported the date this was done in the review. We also searched meeting abstracts via the HPB journal (onlinelibrary.wiley.com/journal/10.1111/(ISSN)1477-2574; accessed 9 January 2021) and the Conference Proceedings Citation Index to explore further relevant clinical studies.
Grey literature databases
We searched the following grey literature databases (accessed 9 January 2021).
-
OpenGrey (opengrey.eu/)
-
PsycEXTRA (www.apa.org/pubs/databases/psycextra/index)
Clinical trials registers and trial result registers
We conducted a search of clinical trial registers/trial result registers (accessed 9 January 2021).
-
ClinicalTrials.gov (www.clinicaltrials.gov/)
-
World Health Organization International Clinical Trials Registry Platform Search Portal (apps.who.int/trialsearch/)
-
Chinese Clinical Trial Registry (www.chictr.org.cn/enindex.aspx)
Data collection and analysis
Selection of studies
Two review authors (HH, Zhuyin Li) independently screened titles and abstracts for inclusion. We coded all of the potential studies we identified as a result of the search as either 'retrieve' (eligible, potentially eligible, or unclear) or 'do not retrieve'. We retrieved the full texts of potentially eligible studies, and two review authors (HH, Zhuyin Li) independently screened the full texts, identified studies for inclusion, and identified and recorded the reasons for exclusion of the ineligible studies. We resolved any disagreement through discussion or, if required, we consulted a third review author (YD). We identified and excluded duplicates and collated multiple reports of the same study, so that each study rather than each report was the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram (Page 2021), and characteristics of excluded studies table.
Data extraction and management
We used a standardized data collection form for study characteristics and outcome data, which was piloted on at least one study included in the review. Two review authors (ZZ, Zuojin Liu) independently extracted the following information from the included studies.
-
Methods: study design, total study duration and run‐in (the time period before entering a clinical trial, in which participants may discontinue or begin treatment), number of study centers and location, study setting, withdrawals, date of study
-
Participants: number (N), mean age, age range, sex, severity of condition, diagnostic criteria, inclusion criteria, exclusion criteria
-
Interventions: intervention, comparison
-
Outcomes: primary and secondary outcomes specified and collected, time points reported
-
Notes: funding for study, notable conflicts of interest of study authors
Two review authors (ZZ, Zuojin Liu) independently extracted outcome data from the included studies. We noted in the characteristics of included studies table if the study authors reported outcome data in an unusable way. We resolved disagreements by consensus or by involving a third review author (YD). One review author (YD) copied across the data from the data collection form into the Review Manager 5 file (Review Manager 2020). We double‐checked that the data were entered correctly by comparing the study reports with how the data were presented in the systematic review. A second review author spot‐checked study characteristics for accuracy against the trial report.
Assessment of risk of bias in included studies
Two review authors (YC, JG) independently assessed the risk of bias for each study using the Cochrane RoB 1 tool for randomized studies, outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We resolved any disagreements by discussion or by involving a third review author (YD). We assessed the risk of bias for the following domains.
-
Random sequence generation
-
Allocation concealment
-
Blinding of participants and personnel
-
Blinding of outcome assessment
-
Incomplete outcome data
-
Selective outcome reporting
-
Other bias
We judged each study to be at high, low or unclear risk of bias for each domain. We provided a quote from the study report and justification for our judgment in the risk of bias table. We summarized the risk of bias judgments across studies for each of the domains listed. We considered blinding separately for different key outcomes where necessary (e.g. for unblinded outcome assessment, risk of bias for postoperative mortality may be very different than for a person‐reported pain scale). Where information on risk of bias related to unpublished data or correspondence with a study author, we noted this in the risk of bias table. If we judged a trial to be at low risk of bias in all domains, we considered it to be at low risk of bias overall. Otherwise, we considered studies to be at high risk of bias (e.g. studies with unclear or high risk of bias for one or more domains). We resolved any difference in opinion by discussion.
When considering treatment effects, we took into account the risk of bias for the studies that contributed to that outcome, as part of the GRADE methodology (Guyatt 2008).
Assessment of bias in conducting the systematic review
We conducted the review according to a protocol (Cheng 2019), and reported any deviations from that protocol in the Differences between protocol and review section of this systematic review.
Measures of treatment effect
We analyzed dichotomous data as risk ratio (RR) and continuous data as mean difference (MD). We ensured that higher scores for continuous outcomes had the same meaning for the particular outcome; we also explained the direction of effect to the reader and reported where the directions were reversed, if this was necessary.
We undertook meta‐analyses only where this was meaningful, that is, if the treatments, participants, and the underlying clinical question were similar enough for pooling to make them meaningful.
A common way that trialists indicated they had skewed data was by reporting medians and interquartile ranges. When we encountered this, we noted that the data might be skewed and considered the implication of this. If the data were skewed, we did not pool them with data from studies that reported the means and standard deviations (SDs), but instead provided a narrative summary. The length of hospital stay might not follow a normal distribution. Therefore, we performed analysis using MDs with caution.
Where multiple trial arms were reported in a single trial, we included only the relevant arms. If two comparisons (e.g. drug A versus placebo and drug B versus placebo) had to be entered into the same meta‐analysis, we halved the control group to avoid double counting.
Unit of analysis issues
The unit of analysis was individual participants undergoing pancreatoduodenectomy. We did not find any cross‐over or cluster‐RCTs for this review.
Dealing with missing data
We contacted investigators or study sponsors to verify key study characteristics and obtain missing numerical outcome data as indicated (e.g. when a study was identified as abstract only).
Assessment of heterogeneity
We intended to describe heterogeneity using the Chi2 test (Higgins 2021). We considered a P value < 0.10 to indicate significant heterogeneity. We also used the I² statistic to measure heterogeneity among studies in each analysis (0% to 40%: might not be important; 30% to 60%: may represent moderate heterogeneity; 50% to 90%: may represent substantial heterogeneity; 75% to 100%: considerable heterogeneity; Higgins 2003). If we identified substantial or considerable heterogeneity, we explored it by prespecified subgroup analysis, and we interpreted summary effect measures with caution.
Assessment of reporting biases
If we were able to pool more than 10 studies, we created and examined a funnel plot to explore possible publication biases. We used visual asymmetry in the funnel plot to determine reporting biases. We also used Egger's test to determine the statistical significance of the reporting bias (Egger 1997). We considered a P value < 0.05 to indicate statistically significant reporting bias.
Data synthesis
We performed the meta‐analysis using Review Manager 5 (Review Manager 2020). We used the random‐effects model by default. To test the robustness of our findings, we conducted sensitivity analyses using the fixed‐effect model. In the case of divergence between the two models, we present both results; otherwise, we present results only from the random‐effects model.
Subgroup analysis and investigation of heterogeneity
We planned to conduct the following subgroup analyses.
-
Pancreatoduodenectomy performed via laparotomy versus laparoscopy
-
Participants at high risk of POPF (soft pancreas) versus participants at low risk of POPF (firm pancreas)
We used the following outcomes in subgroup analyses.
-
Rate of POPF
-
Postoperative mortality
We used the formal Chi² test for subgroup differences to test for subgroup interactions.
Sensitivity analysis
We planned to perform the following sensitivity analyses for all outcomes, defined a priori, to assess the robustness of our conclusions. The sensitivity analyses involved:
-
changing between fixed‐effect and random‐effects models;
-
changing between worst‐case scenario analysis (the events happened in the experimental group but did not happen in the control group for missing participants) and best‐case scenario analysis (the events happened in the control group but did not happen in the experimental group for missing participants) for missing data;
-
excluding studies in which the mean or SD, or both, were imputed;
-
excluding studies that did not use the International Study Group of Pancreatic Fistula (ISGPF) definition of POPF (Bassi 2005; Bassi 2017).
We performed a posthoc sensitivity analysis by including studies that did or did not use the ISGPF definition of POPF (Bassi 2005; Bassi 2017), according to the editor's suggestion.
Reaching conclusions
We based our conclusions on findings from the quantitative or narrative synthesis of studies included in this review. The 'Implications for research' section provides the reader with a clear sense of the remaining uncertainties in the field and the direction of focus required of future research.
Summary of findings and assessment of the certainty of the evidence
We created a summary of findings table, including the following outcomes: rate of POPF, postoperative mortality, adverse events, rate of surgical reintervention, rate of postoperative bleeding, overall rate of surgical complications, length of hospital stay, and quality of life. We used the five GRADE considerations (risk of bias, consistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of evidence based on the studies that contributed data to the meta‐analyses for each outcome, classifying the certainty as high, moderate, low, or very low. We used the methods and recommendations described in Chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021), and GRADEpro GDT software (GRADEpro GDT). We justified all decisions to downgrade the certainty of the included studies in the footnotes and provided comments where necessary to aid the reader's understanding of the review. Two review authors (HH, Zhuyin Li) independently justified all decisions to downgrade the certainty of studies. We considered whether there was additional outcome information that could not be incorporated into the meta‐analyses; we noted this in the comments, and stated whether it supported or contradicted the information from the meta‐analyses.
We defined the certainty of levels of evidence as 'high', 'moderate', 'low', or 'very low' as follows.
-
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
-
Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
-
Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect
-
Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
Results of the search
We identified 1917 records through electronic searches of the Cochrane Library (249 records), MEDLINE (Ovid) (364 records), Embase (Ovid) (742 records), and Science Citation Index Expanded (Web of Science) (562 records). We identified two records through scanning the reference lists of the identified RCTs. We excluded 355 duplicates and 1541 clearly irrelevant records after reading the titles and abstracts. We retrieved the remaining 23 records for further assessment. We excluded nine studies for the reasons listed in the Characteristics of excluded studies table, one study awaiting classification, and two ongoing studies. In total, 11 RCTs fulfilled the inclusion criteria for this review. The study flow diagram is shown in Figure 1.
Study flow diagram.
Included studies
We included 10 RCTs comparing duct‐to‐mucosa pancreaticojejunostomy with any other type of pancreaticojejunostomy (invagination pancreaticojejunostomy) and all provided data for the analyses. We included one RCT comparing one type of duct‐to‐mucosa pancreaticojejunostomy (modified Blumgart technique) with a different type of duct‐to‐mucosa pancreaticojejunostomy (traditional interrupted technique). Details of the studies are shown in the Characteristics of included studies table.
Duct‐to‐mucosa versus any other type of pancreaticojejunostomy (invagination pancreaticojejunostomy)
We included 10 RCTs published between 1996 and 2018 comparing duct‐to‐mucosa pancreaticojejunostomy with invagination pancreaticojejunostomy (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). Only one RCT was a dual‐center study (Berger 2009); the other nine RCTs were single‐center studies. The RCTs were conducted in the USA (N = 1; Berger 2009), China (N = 4; Bai 2016; Chou 1996; Han 2009; Xu 2015), Egypt (N = 1; El Nakeeb 2015), Germany (N = 1; Langrehr 2005), India (N = 1; Singh 2018), Italy (N = 1; Bassi 2003), and Japan (N = 1; Senda 2018). The number of surgeons performing the pancreaticojejunostomy was reported in eight RCTs, and it ranged from one to eight in those studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Senda 2018; Singh 2018; Xu 2015). The experience of the surgeons with the two types of pancreaticojejunostomy was reported in only four RCTs (Bai 2016; Berger 2009; Senda 2018; Xu 2015). Three of those four RCTs involved experienced surgeons (Bai 2016; Senda 2018; Xu 2015), while one RCT reported relatively inexperienced surgeons performing one of the techniques (Berger 2009).
A total of 1472 participants were randomized to either the duct‐to‐mucosa group (N = 732) or the invagination group (N = 740). The sample size calculation was described in six of the 10 RCTs (Bai 2016; Berger 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). All RCTs reported the mean age and sex distribution of the included participants. The mean age of the participants ranged from 54.0 years to 68.0 years. The mean proportion of females varied between 31.2% and 49.7%.
The diagnoses of the participants included were given in seven studies (Bai 2016; Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Senda 2018; Singh 2018). The diagnoses were as follows: pancreatic cancer (34.4%), biliary cancer (9.5%), ampullary cancer (32.3%), chronic pancreatitis (3.4%), and other diseases (20.4%). The remaining three RCTs did not report the proportions of the different diagnoses (Berger 2009; El Nakeeb 2015; Xu 2015).
Six RCTs included both participants with soft pancreatic texture (ranging from 46.6% to 64.6%) and participants with firm pancreatic texture (ranging from 35.4% to 53.4%) (Bai 2016; Berger 2009; El Nakeeb 2015; Senda 2018; Singh 2018; Xu 2015). Two RCTs included only participants with soft pancreatic texture (Bassi 2003; Han 2009). The remaining two RCTs did not report pancreatic texture (Chou 1996; Langrehr 2005).
Pancreatic duct size was reported in four studies (Bai 2016; Berger 2009; El Nakeeb 2015; Singh 2018). The mean pancreatic duct diameter was similar between the duct‐to‐mucosa group (ranging from 3 mm to 4.3 mm) and the invagination group (ranging from 3 mm to 4.1 mm). The remaining six RCTs did not report pancreatic duct size (Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Senda 2018; Xu 2015).
Placement of prophylactic pancreatic duct stents was reported in seven RCTs (Bai 2016; Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Senda 2018; Singh 2018). Four of the seven RCTs used external stents (Bassi 2003; Chou 1996; Langrehr 2005; Senda 2018), two RCTs used internal stents (Bai 2016; Han 2009), and one RCT used both external and internal stents (Singh 2018). The proportion of participants using pancreatic duct stents ranged from 21.1% to 100% in seven RCTs (Bai 2016; Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Senda 2018; Singh 2018). The remaining three RCTs did not use any stents when performing pancreaticojejunostomy (Berger 2009; El Nakeeb 2015; Xu 2015).
Administration of somatostatin or its analogues was reported in four RCTs, with the proportion of participants using prophylactic somatostatin or its analogues ranging from 35.6% to 100% (Bassi 2003; Langrehr 2005; Singh 2018; Xu 2015). Four RCTs did not use somatostatin or its analogues (Bai 2016; Berger 2009; Han 2009; Senda 2018). The remaining two RCTs did not mention the use of somatostatin or its analogues (Chou 1996; El Nakeeb 2015).
The outcomes measured were: rate of POPF, grade of POPF, postoperative mortality, overall rate of surgical complications, rate of surgical reintervention, rate of postoperative bleeding, and length of hospital stay. Seven of the 10 RCTs reported the rate of POPF according to the 2005 ISGPF definition of POPF (a drain output of any measurable volume of fluid on or after postoperative day 3 with an amylase content greater than 3 times the serum amylase activity; Bassi 2005) (Bai 2016; Berger 2009; El Nakeeb 2015; Han 2009; Senda 2018; Singh 2018; Xu 2015). The proportion of participants with grade B POPF was 8.3%, and the proportion of participants with grade C POPF was 1.7%. None of the included RCTs used the 2016 ISGPS definition of POPF (Bassi 2017). The remaining three studies used neither the 2005 ISGPF definition nor the 2016 ISGPS definition of POPF (Bassi 2003; Chou 1996; Langrehr 2005). Bassi 2003 defined POPF as "output > 30 mL/24 hours; rich in amylase content for at least 7 days from postoperative day 4, confirmed by fistulography". Chou 1996 defined POPF as "persistent drainage of 50 mL or more of amylase‐rich fluid a day for more than two weeks". Langrehr 2005 defined POPF as "drainage fluid with elevated amylase and lipase levels from postoperative day 5 on > 1000 U/L and beyond postoperative day 10, clinical symptoms (pain, fever, etc.)".
Four RCTs were funded by non‐commercial grants (e.g. university grants, national cancer control programs, charitable funding; Bai 2016; Bassi 2003; Han 2009; Xu 2015). One RCT did not receive any funding (El Nakeeb 2015). The other five RCTs gave no information about funding sources (Berger 2009; Chou 1996; Langrehr 2005; Senda 2018; Singh 2018).
Four RCTs declared no conflict of interest (Bai 2016; Berger 2009; El Nakeeb 2015; Senda 2018). The other six RCTs did not report whether there were conflicts of interest (Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Singh 2018; Xu 2015).
One type of duct‐to‐mucosa pancreaticojejunostomy versus a different type of duct‐to‐mucosa pancreaticojejunostomy
We included one RCT comparing one type of duct‐to‐mucosa pancreaticojejunostomy (modified Blumgart technique) with a different type of duct‐to‐mucosa pancreaticojejunostomy (traditional interrupted technique).
The single‐center study was conducted in Japan and published in 2019 (Hirono 2019). The number and experience of the surgeons performing pancreaticojejunostomy were not reported in this study. A total of 224 participants were randomized to either the modified Blumgart technique (N = 112) or the traditional interrupted group (N = 112). The sample size calculation was described in this RCT. The ages of the participants ranged from 24 years to 90 years and the proportion of females was 42.4%. The diagnoses of the participants included were as follows: pancreatic cancer (36.2%), biliary cancer (29%), and other diseases (34.8%). This RCT included participants with soft pancreatic texture (56.7%) and participants with firm pancreatic texture (43.3%). The mean pancreatic duct diameter was the same between the modified Blumgart group (median 3 mm) and the traditional interrupted group (median 3 mm). The placement of prophylactic pancreatic duct stents (internal stents) was applied in this RCT. Somatostatin or its analogues were not administered in this RCT.
The outcomes measured in this RCT were: rate of POPF, grade of POPF, postoperative mortality, overall rate of surgical complications, rate of surgical reintervention, rate of postoperative bleeding, and length of hospital stay. This RCT reported the rate of POPF according to the 2016 ISGPS definition of POPF (a drain output of any measurable volume of fluid with an amylase level > 3 times the upper limit of institutional normal serum amylase activity, associated with a clinically relevant development/condition related directly to the POPF; Bassi 2017). The proportion of participants with grade B POPF was 8.5%. There were no participants with grade C POPF. This RCT declared no conflicts of interest but did not provide any information about funding sources.
Excluded studies
We excluded nine studies. The details of these studies are listed in the Characteristics of excluded studies table. We excluded two studies because they compared invagination pancreaticojejunostomy with binding pancreaticojejunostomy or with ligation of the pancreatic duct (Peng 2007; Reissman 1995). The remaining excluded studies were identified as non‐randomized studies (Batignani 2005; Casadei 2020; Lee 2018; Peng 2003; Wei 2015; Wu 2019; Zhang 2013).
Studies awaiting classification
One study was reported in a conference abstract (Velineni 2019). That study did not report further information on the study design, inclusion criteria, or exclusion criteria. We received no reply from the authors on email contact. We have therefore assessed this study as awaiting classification (Characteristics of studies awaiting classification).
Ongoing studies
NCT03600584: participants (380) scheduled to undergo open pancreaticoduodenectomy will be randomized to one‐layer duct‐to‐mucosa pancreaticojejunostomy or invagination pancreaticojejunostomy. This trial is currently recruiting participants and is being performed in China; it was initiated in July 2018. The primary outcome is the rate of POPF. The secondary outcomes are anastomosis time, rate of delayed gastric emptying, rate of postoperative bleeding, rate of chyle leakage, overall rate of surgical complications, mortality, rate of surgical reintervention, readmission rate, and duration of postoperative hospital stay (Characteristics of ongoing studies).
Pan 2016: participants (114) scheduled to undergo pancreaticoduodenectomy will be randomized to one‐layer duct‐to‐mucosa pancreaticojejunostomy or two‐layer duct‐to‐mucosa pancreaticojejunostomy. This trial is currently recruiting participants and is being performed in China; it was initiated in July 2015. The primary outcome is the rate of POPF. The secondary outcomes are duration of postoperative hospital stay, anastomosis time, rate of surgical reintervention, overall rate of surgical complications, mortality, rate of biliary leakage, and blood transfusion (Characteristics of ongoing studies).
Risk of bias in included studies
We judged all 11 studies to be at high risk of bias (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Hirono 2019; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). For graphical representations of the risk of bias, see Figure 2 and Figure 3.
Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgments about each risk of bias item for each included study.
Allocation
Random sequence generation
Participants were randomized using computer‐generated numbers in five studies (Bai 2016; Berger 2009; Hirono 2019; Singh 2018; Xu 2015), and using random number tables in two studies (Bassi 2003; Han 2009); we judged random sequence generation to be at low risk of bias in seven studies (Bai 2016; Bassi 2003; Berger 2009; Han 2009; Hirono 2019; Singh 2018; Xu 2015). No information was provided regarding random sequence generation, and we received no responses from the authors in the remaining four studies (Chou 1996; El Nakeeb 2015; Langrehr 2005; Senda 2018); we judged random sequence generation to be at unclear risk of bias in those four studies.
Allocation concealment
The treatment allocation was concealed using sealed opaque envelopes in four studies (Bai 2016; Berger 2009; El Nakeeb 2015; Singh 2018), and central allocation in two studies (Senda 2018; Xu 2015); we therefore judged allocation concealment to be at low risk of bias in those six studies. No information was provided regarding allocation concealment, and we received no responses from the authors in the remaining five studies (Bassi 2003; Chou 1996; Han 2009; Hirono 2019; Langrehr 2005); we therefore judged allocation concealment to be at unclear risk of bias in those five studies.
Blinding
Blinding of participants and personnel
Due to the nature of the interventions, the surgeons could not be blinded to group allocation; we therefore judged blinding of participants and personnel to be at high risk of bias in all studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Hirono 2019; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015).
Blinding of outcome assessment
Two studies utilized appropriate strategies for blinding outcome assessors (Bai 2016; El Nakeeb 2015); we therefore judged blinding of outcome assessment to be at low risk of bias in those two studies. No information was provided regarding blinding of outcome assessment, and we received no responses from the authors in five studies (Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Senda 2018); we therefore judged blinding of outcome assessment to be at unclear risk of bias in those five studies. The remaining four studies mentioned "open label" in the protocols (Berger 2009; Hirono 2019; Singh 2018; Xu 2015); we therefore judged blinding of outcome assessment to be at high risk of bias in those four studies.
Incomplete outcome data
There were no post‐randomization dropouts in eight of the studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; Han 2009; Langrehr 2005; Senda 2018; Xu 2015). Although there were seven dropouts (6.5%) in one trial, the data were analyzed on an intention‐to‐treat basis (El Nakeeb 2015). There was one dropout (0.5%) in another trial (Singh 2018), and the trial reported the outcome of this patient. The reason for the dropout is reported in the Characteristics of included studies table. We therefore judged incomplete outcome data to be at low risk of bias in these 10 studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). There were 14 dropouts (6.3%) in the remaining trial, and the data were not analyzed on an intention‐to‐treat basis (Hirono 2019). The reasons for the dropouts are reported in the Characteristics of included studies table. We therefore judged incomplete outcome data to be at high risk of bias in this trial (Hirono 2019).
Selective reporting
The study protocols were available for six studies and all prespecified outcomes were reported in these studies (Bai 2016; Berger 2009; El Nakeeb 2015; Hirono 2019; Senda 2018; Singh 2018). The study protocols were not available for four studies, and all expected outcomes were reported fully (Bassi 2003; Chou 1996; Han 2009; Langrehr 2005). We therefore judged selective reporting to be at low risk of bias in these 10 studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Hirono 2019; Langrehr 2005; Senda 2018; Singh 2018). In the remaining trial (Xu 2015), two of the study's prespecified outcomes in the study protocol (anastomosis time and overall rate of surgical complications) were not reported; we therefore judged selective reporting to be at high risk of bias in this trial (Xu 2015).
Other potential sources of bias
No baseline imbalances were observed; therefore we judged baseline imbalance to be at low risk of bias in all studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Hirono 2019; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015).
Effects of interventions
See: Summary of findings 1 Duct‐to‐mucosa compared to invagination pancreaticojejunostomy for the prevention of postoperative pancreatic fistula following pancreaticoduodenectomy; Summary of findings 2 Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique for the prevention of postoperative pancreatic fistula following pancreaticoduodenectomy
Duct‐to‐mucosa versus any other type of pancreaticojejunostomy
We included 10 RCTs (1472 participants) comparing duct‐to‐mucosa pancreaticojejunostomy with invagination pancreaticojejunostomy (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). Participants were randomized to the duct‐to‐mucosa group (732 participants) and the invagination group (740 participants) (summary of findings Table 1). We did not identify any RCTs comparing duct‐to‐mucosa pancreaticojejunostomy with any other type of pancreaticojejunostomy.
Rate of POPF (grade B or C defined according to the 2005 ISGPF definition)
Ten studies reported the rate of POPF (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). Three of these studies reported the rate of POPF using neither the 2005 ISGPF nor the 2016 ISGPS definition; therefore, to reduce bias, we did not include these three studies in the quantitative synthesis (meta‐analysis) (Bassi 2003; Chou 1996; Langrehr 2005). The remaining seven studies included 1122 participants (557 duct‐to‐mucosa, 565 invagination) and reported the rate of POPF according to the 2005 ISGPF definition (Bai 2016; Berger 2009; El Nakeeb 2015; Han 2009; Senda 2018; Singh 2018; Xu 2015). Overall, there were 112 cases of POPF (grade B or C ): 67 (56 grade B, 11 grade C) in the duct‐to‐mucosa group and 45 (37 grade B, 8 grade C) in the invagination group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of POPF (grade B or C defined according to the 2005 ISGPF definition) (risk ratio (RR) 1.45, 95% confidence interval (CI) 0.64 to 3.26; P = 0.37; test for heterogeneity, P = 0.004, I2 = 68%; 7 studies, 1122 participants; very low‐certainty evidence; Analysis 1.1).
With regard to grades of POPF, the evidence is also very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on either the rate of grade B POPF (RR 1.41, 95% CI 0.63 to 3.15; test for heterogeneity, P = 0.02, I2 = 62%; 7 studies, 1122 participants; very low‐certainty evidence; Analysis 1.1) or the rate of grade C POPF (RR 1.45, 95% CI 0.58 to 3.61; test for heterogeneity, P = 0.64, I2 = 0%; 7 studies, 1122 participants; very low‐certainty evidence; Analysis 1.1).
We performed a posthoc sensitivity analysis in which we included three studies that did not use the ISGPF definition of POPF (Bassi 2003; Chou 1996; Langrehr 2005). Overall, there were 145 cases of POPF: 80 in the duct‐to‐mucosa group and 65 in the invagination group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of POPF (RR 1.13, 95% CI 0.60 to 2.15; test for heterogeneity, P = 0.003, I2 = 64%; 10 studies, 1472 participants; very low‐certainty evidence; Analysis 2.1).
Postoperative mortality
All included studies (1472 participants; 732 duct‐to‐mucosa, 740 invagination) reported postoperative mortality. There were 15 deaths in the duct‐to‐mucosa group and 19 deaths in the invagination group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on postoperative mortality compared with invagination pancreaticojejunostomy (RR 0.77, 95% CI 0.39 to 1.49; P = 0.43; test for heterogeneity, P = 0.77, I2 = 0%; 10 studies, 1472 participants; very low‐certainty evidence; Analysis 1.2).
Adverse events
None of the studies reported this outcome.
Rate of surgical reintervention
All included studies (1472 participants; 732 duct‐to‐mucosa, 740 invagination) reported the rate of surgical reintervention. Overall, 50 participants needed surgical reintervention: 25 in the duct‐to‐mucosa group and 25 in the invagination group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of surgical reintervention (RR 1.12, 95% CI 0.65 to 1.95; P = 0.68; test for heterogeneity, P = 0.63, I2 = 0%; 10 studies, 1472 participants; very low‐certainty evidence; Analysis 1.3).
Rate of postoperative bleeding
Nine studies (1275 participants; 635 duct‐to‐mucosa, 640 invagination) reported the rate of postoperative bleeding (Bai 2016; Bassi 2003; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015); another study did not report this outcome (Berger 2009). Overall, there were 56 cases of postoperative bleeding: 26 in the duct‐to‐mucosa group and 30 in the invagination group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of postoperative bleeding (RR 0.85, 95% CI 0.51 to 1.42; P = 0.54; test for heterogeneity, P = 0.95, I2 = 0%; 9 studies, 1275 participants; very low‐certainty evidence; Analysis 1.4).
Overall rate of surgical complications
Five studies (750 participants; 372 duct‐to‐mucosa, 378 invagination) reported the overall rate of surgical complications (Bai 2016; Berger 2009; El Nakeeb 2015; Senda 2018; Singh 2018). The other five studies did not report this outcome (Bassi 2003; Chou 1996; Han 2009; Langrehr 2005; Xu 2015). Overall, there were 384 participants suffering surgical complications: 201 in the duct‐to‐mucosa group and 183 in the invagination group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the overall rate of surgical complications (RR 1.12, 95% CI 0.92 to 1.36; P = 0.24; test for heterogeneity, P = 0.12, I2 = 45%; 5 studies, 750 participants; very low‐certainty evidence; Analysis 1.5).
Length of hospital stay
All included studies (1471 participants; 732 duct‐to‐mucosa, 739 invagination) reported length of hospital stay. Four studies reported the mean and standard deviation (SD) (Chou 1996; Han 2009; Singh 2018; Xu 2015). The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy on the length of hospital stay compared with invagination pancreaticojejunostomy (mean difference (MD) ‐0.41 days, 95% CI ‐1.87 to 1.04; P = 0.58; test for heterogeneity, P = 0.40, I2 = 0%; 4 studies, 658 participants; very low‐certainty evidence Analysis 1.6). Six studies with skewed data reported median and range values which were not suitable for pooling (Bai 2016; Bassi 2003; Berger 2009; El Nakeeb 2015; Langrehr 2005; Senda 2018). The findings of these six studies (813 participants) were mixed. Bai 2016 reported that the duct‐to‐mucosa group was associated with less time in the hospital than the invagination group (13 days versus 15 days), but Senda 2018 reported that the duct‐to‐mucosa group was associated with more time in the hospital (24 days versus 19 days) than the invagination group. The remaining four studies reported no difference in the length of hospital stay between groups (Bassi 2003; Berger 2009; El Nakeeb 2015; Langrehr 2005; Analysis 1.7).
Quality of life
None of the studies reported this outcome.
One type of duct‐to‐mucosa pancreaticojejunostomy versus a different type of duct‐to‐mucosa pancreaticojejunostomy
We included one RCT comparing one type of duct‐to‐mucosa pancreaticojejunostomy (modified Blumgart technique) to a different type of duct‐to‐mucosa pancreaticojejunostomy (traditional interrupted technique). Half of the participants (112) were randomized to the modified Blumgart group and half (112 participants) to the traditional interrupted group (summary of findings Table 2).
Rate of POPF (grade B or C defined according to the 2016 ISGPS definition)
One study (210 participants; 107 modified Blumgart technique, 103 traditional interrupted technique) reported the rate of POPF according to the 2016 ISGPS definition (Hirono 2019). Overall, there were 18 cases of POPF (18 grade B, no grade C): 11 (11 grade B, no grade C) in the modified Blumgart group and seven (7 grade B, no grade C) in the traditional interrupted group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of POPF (grade B or C defined according to the 2016 ISGPS definition) (RR 1.51, 95% CI 0.61 to 3.75; 1 study, 210 participants; very low‐certainty evidence; Analysis 3.1).
With regard to grades of POPF, there were no participants with grade C POPF in either group. The evidence is also very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of grade B POPF (RR 1.51, 95% CI 0.61 to 3.75; P = 0.37; 1 study, 210 participants; very low‐certainty evidence; Analysis 3.1).
Postoperative mortality
One study (210 participants; 107 modified Blumgart technique, 103 traditional interrupted technique) reported postoperative mortality (Hirono 2019). There were no deaths in either group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on postoperative mortality.
Adverse events
The study did not report this outcome.
Rate of surgical reintervention
One study (210 participants; 107 modified Blumgart technique, 103 traditional interrupted technique) reported the rate of surgical reintervention (Hirono 2019). Overall, three participants needed surgical reintervention: two in the modified Blumgart group and one in the traditional interrupted group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of surgical reintervention (RR 1.93, 95% CI 0.18 to 20.91; 1 study, 210 participants; very low‐certainty evidence; Analysis 3.3).
Rate of postoperative bleeding
One study (210 participants; 107 modified Blumgart technique, 103 traditional interrupted technique) reported the rate of postoperative bleeding (Hirono 2019). Overall, there was only one case of postoperative bleeding, and it occurred in the modified Blumgart group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of postoperative bleeding (RR 2.89, 95% CI 0.12 to 70.11; 1 study, 210 participants; very low‐certainty evidence; Analysis 3.4).
Overall rate of surgical complications
One study (210 participants; 107 modified Blumgart technique, 103 traditional interrupted technique) reported the overall rate of surgical complications (Hirono 2019). Overall, 90 participants suffered surgical complications: 48 in the modified Blumgart group and 42 in the traditional interrupted group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the overall rate of surgical complications (RR 1.10, 95% CI 0.80 to 1.51; 1 study, 210 participants; very low‐certainty evidence; Analysis 3.5).
Length of hospital stay
One study (210 participants; 107 modified Blumgart technique, 103 traditional interrupted technique) reported the length of hospital stay (Hirono 2019). This study reported the median and range values for length of hospital stay; however, the data were skewed, and the values were not suitable for pooling (Hirono 2019). The median length of hospital stay in the modified Blumgart group was 15 days (range 8‐52 days). The median length of hospital stay in the traditional interrupted group was 15 days (range 6‐44 days) (Analysis 3.6). The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the length of hospital stay.
Quality of life
The study did not report this outcome.
Subgroup analysis
We performed a planned subgroup analysis of participants at high risk of POPF (soft pancreas) versus participants at low risk of POPF (firm pancreas) (Analysis 4.1; Analysis 4.2). These subgroup analyses did not reveal any differences in the primary outcomes considered in this review.
We were unable to perform a planned subgroup analysis based on different types of pancreatoduodenectomy (e.g. laparotomy, laparoscopy) because none of the studies performed laparoscopic pancreatoduodenectomy.
Sensitivity analysis
We performed the following sensitivity analyses.
-
Changing between fixed‐effect and random‐effects models
-
Changing between worst‐case scenario analysis and best‐case scenario analysis for missing data
-
Posthoc sensitivity analysis by including studies that did or did not use the International Study Group of Pancreatic Fistula (ISGPF) definition of POPF (Bassi 2005; Bassi 2017)
For the comparison of duct‐to‐mucosa pancreaticojejunostomy with invagination pancreaticojejunostomy, we observed no change in the results by changing between fixed‐effect and random‐effects models, except for the outcome 'rate of POPF (grade B or C defined according to the 2005 ISGPF definition)' (Analysis 5.1). The rate of POPF (grade B or C defined according to the 2005 ISGPF definition) was significantly lower in the invagination group than in the duct‐to‐mucosa group when the fixed‐effect model was used (RR 1.50, 95% CI 1.05 to 2.14). However, when the random‐effects model was used, there was no significant difference in the rate of POPF (grade B or C defined according to the 2005 ISGPF definition) between the groups (RR 1.44, 95% CI 0.64 to 3.26). We observed no change in the results obtained using worst‐case and best‐case scenario analysis for missing data (Table 1), or when we included studies that did or did not use the ISGPF definition of POPF (Analysis 2.1).
| Comparisons | Outcomes | Main analysis (95% CI) | Sensitivity analysis using fixed‐effect model (95% CI) | Sensitivity analysis by changing between worst‐case scenario analysis and best‐case scenario analysis for missing data | Posthoc sensitivity analysis by including studies that did or did not use the ISGPF definition of POPF | |
|---|---|---|---|---|---|---|
| Worst/best‐case (95% CI) | Best/worst‐case (95% CI) | |||||
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Rate of POPF (grade B or C defined according to the 2005 ISGPF definition) | RR 1.44 (0.64 to 3.26) | RR 1.50 (1.05 to 2.15) | RR 1.44 (0.64 to 3.26) | RR 1.44 (0.64 to 3.26) | RR 1.13 (0.60 to 2.15) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Postoperative mortality | RR 0.77 (0.39 to 1.49) | RR 0.81 (0.44 to 1.52) | RR 0.77 (0.39 to 1.49) | RR 0.77 (0.39 to 1.49) | RR 0.77 (0.39 to 1.49) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Adverse events | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Rate of surgical reintervention | RR 1.12 (0.65 to 1.95) | RR 1.01 (0.61 to 1.70) | RR 1.12 (0.65 to 1.95) | RR 1.12 (0.65 to 1.95) | RR 1.12 (0.65 to 1.95) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Rate of postoperative bleeding | RR 0.85 (0.51 to 1.42) | RR 0.87 (0.53 to 1.44) | RR 0.85 (0.51 to 1.42) | RR 0.85 (0.51 to 1.42) | RR 0.85 (0.51 to 1.42) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Overall rate of surgical complications | RR 1.12 (0.92 to 1.36) | RR 1.11 (0.97 to 1.28) | RR 1.12 (0.93 to 1.36) | RR 1.12 (0.93 to 1.36) | RR 1.12 (0.92 to 1.36) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Length of hospital stay | MD ‐0.41 (‐1.87 to 1.04) | MD ‐0.41 (‐1.87 to 1.04) | ‐ | ‐ | MD ‐0.41 (‐1.87 to 1.04) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Quality of life | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Rate of POPF (grade B or C defined according to the 2016 ISGPS definition) | RR 1.51 (0.61 to 3.75) | RR 1.51 (0.61 to 3.75) | RR 2.29 (0.98 to 5.34) | RR 0.69 (0.33 to 1.41) | RR 1.51 (0.61 to 3.75) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Postoperative mortality | No events | No events | RR 11.00 (0.62 to 196.60) | RR 0.05 (0.00 to 0.89) | No events |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Adverse events | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Rate of surgical reintervention | RR 1.93 (0.18 to 20.91) | RR 1.93 (0.18 to 20.91) | RR 7.00 (0.88 to 55.97) | RR 0.20 (0.04 to 0.89) | RR 1.93 (0.18 to 20.91) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Rate of postoperative bleeding | RR 2.89 (0.12 to 70.11) | RR 2.89 (0.12 to 70.11) | RR 13.00 (0.74 to 228.05) | RR 0.11 (0.01 to 0.86) | RR 2.89 (0.12 to 70.11) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Overall rate of surgical complications | RR 1.10 (0.80 to 1.51) | RR 1.10 (0.80 to 1.51) | RR 1.26 (0.93 to 1.72) | RR 0.94 (0.70 to 1.26) | RR 1.10 (0.80 to 1.51) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Length of hospital stay | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Quality of life | ‐ | ‐ | ‐ | ‐ | ‐ |
CI: confidence interval ; RR: risk ratio; MD: mean difference; POPF: postoperative pancreatic fistula; ISGPF: International Study Group on Pancreatic Fistula; ISGPS: International Study Group on Pancreatic Surgery; ‐: not available
For the comparison of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, we observed no change in the results by changing between fixed‐effect and random‐effects models, or including studies that did or did not use the ISGPF definition of POPF (Table 1). We observed no change in the results by changing between worst‐case and best‐case scenario analysis for missing data except for the outcomes 'postoperative mortality', 'rate of surgical reintervention', and 'rate of postoperative bleeding' (Table 1). The differences in the rates of postoperative mortality, surgical reintervention, and postoperative bleeding became statistically significant, favoring the modified Blumgart technique, when best‐case scenario analysis was performed for missing data (Table 1).
For the preceding two comparisons, we did not perform the planned sensitivity analysis by excluding studies in which the mean or SD, or both, were imputed because there were no imputed data. We did not need to perform the planned sensitivity analysis by "excluding studies that did not use International Study Group of Pancreatic Fistula (ISGPF) definition of POPF" because we had already excluded these types of studies when evaluating the rate of POPF.
Discussion
Summary of main results
Evidence from 11 studies, consisting of 1696 participants undergoing pancreaticoduodenectomy, contributed data to the primary outcomes of this review. For the comparison of duct‐to‐mucosa pancreaticojejunostomy with invagination pancreaticojejunostomy (10 studies, 1472 participants), the evidence is very uncertain for the rate of POPF (grade B or C defined according to the 2005 ISGPF definition) and postoperative mortality. None of the studies reported adverse events. For the comparison of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique (1 study, 224 participants), the evidence is very uncertain for the rate of POPF (grade B or C defined according to the 2016 ISGPS definition) and postoperative mortality. The study did not report adverse events.
Overall completeness and applicability of evidence
All studies included people undergoing elective open pancreaticoduodenectomy for various pancreatic and extra‐pancreatic diseases, including pancreatic cancer, ampullary cancer, chronic pancreatitis, and biliary and duodenal malignancies. The majority (71.9%) of the participants had either pancreatic cancer (34.3%) or ampullary cancer (37.6%). None of the studies included participants who underwent laparoscopic pancreaticoduodenectomy. Therefore, the results of this review are only applicable to people undergoing elective open pancreaticoduodenectomy for various pancreatic and extra‐pancreatic diseases, especially for pancreatic cancer and ampullary cancer.
The definition of POPF varied in different studies (see the notes in Characteristics of included studies). The incidence of POPF ranged from 3.9% to 16.7%, according to the different definitions applied in the individual studies (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Hirono 2019; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015). To avoid bias due to differences in the definitions of POPF when evaluating the rate of POPF, we included only the outcome data of studies that used the 2005 ISGPF definition or the 2016 ISGPS definition. The safety of duct‐to‐mucosa pancreaticojejunostomy for people undergoing pancreaticoduodenectomy is another major concern for patients, surgeons, and healthcare funders. In this review, we found that postoperative mortality was less than 5% in both the duct‐to‐mucosa group and the invagination group. There were no deaths in the study that compared duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique with the traditional interrupted technique.
Certainty of the evidence
For the comparison of duct‐to‐mucosa pancreaticojejunostomy with invagination pancreaticojejunostomy, we rated the certainty of evidence for six outcomes, including rate of POPF (grade B or C defined according to the 2005 ISGPF definition), postoperative mortality, rate of surgical reintervention, rate of postoperative bleeding, overall rate of surgical complications, and length of hospital stay.
We graded the certainty of evidence as very low for the rate of POPF (grade B or C defined according to the 2005 ISGPF definition). We downgraded the certainty of evidence by one level for serious risk of bias because seven studies had high risk of bias for blinding of participants and personnel (Bai 2016; Berger 2009; El Nakeeb 2015; Han 2009; Senda 2018; Singh 2018; Xu 2015), and three studies had high risk of bias for blinding of outcome assessment (Berger 2009; Singh 2018; Xu 2015). We downgraded the certainty of evidence by one level for serious unexplained inconsistency (substantial heterogeneity I2 = 68%). We downgraded the certainty of evidence by one level for serious imprecision because of the wide confidence interval that included both potential benefit and potential harm from the intervention. We downgraded the certainty of the evidence by one level for indirectness because of the different numbers of performing surgeons with different levels of experience and different proportions of participants who received pancreatic duct stents and somatostatin or its analogues among the included studies. We downgraded the certainty of evidence by one level for publication bias because the funnel plot was asymmetrical.
We graded the certainty of evidence as very low for postoperative mortality. Due to the nature of the interventions, the performing surgeons could not be blinded in any of the studies. Three studies with high risk of bias for blinding of outcome assessors contributed 42.9% of the weight toward this effect estimate (Berger 2009; Singh 2018; Xu 2015); however, because unblinding of participants, performing surgeons, or outcome assessors is not expected to impact on postoperative mortality, we did not downgrade the certainty of evidence for risk of bias. We downgraded the certainty of evidence by two levels for very serious imprecision because of few events and a wide confidence interval that included both potential benefit and potential harm from the intervention. We downgraded the certainty of the evidence by one level for indirectness because of the different numbers of performing surgeons with different levels of experience and different proportions of participants who received pancreatic duct stents and somatostatin or its analogues among the included studies. We downgraded the certainty of evidence by one level for publication bias because the funnel plot displayed asymmetry (Figure 4) (Egger's test, P = 0.046).
Funnel plot of comparison: 1 Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy, outcome: 1.2 Postoperative mortality.
We graded the certainty of evidence as very low for the rate of surgical reintervention. We downgraded the certainty of evidence by one level for serious risk of bias because 10 studies had high risk of bias for blinding of participants and personnel (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015), and three studies had high risk of bias for blinding of outcome assessment (Berger 2009; Singh 2018; Xu 2015). We downgraded the certainty of evidence by two levels for very serious imprecision because of few events and a wide confidence interval that included both potential benefit and potential harm from the intervention. We downgraded the certainty of the evidence by one level for indirectness because of the different numbers of performing surgeons with different levels of experience and different proportions of participants who received pancreatic duct stents and somatostatin or its analogues among the included studies. We downgraded the certainty of evidence by one level for publication bias because the funnel plot displayed asymmetry (Figure 5) (Egger's test, P = 0.016).
Funnel plot of comparison: 1 Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy, outcome: 1.3 Rate of surgical reintervention.
We graded the certainty of evidence as very low for the rate of postoperative bleeding. We downgraded the certainty of evidence by one level for serious risk of bias because nine studies had high risk of bias for blinding of participants and personnel (Bai 2016; Bassi 2003; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015), and two studies had high risk of bias for blinding of outcome assessment (Singh 2018; Xu 2015). We downgraded the certainty of evidence by two levels for very serious imprecision because of few events and a wide confidence interval that included both potential benefit and potential harm from the intervention. We downgraded the certainty of the evidence by one level for indirectness because of the different numbers of performing surgeons with different levels of experience and different proportions of participants who received pancreatic duct stents and somatostatin or its analogues among the included studies. We downgraded the certainty of evidence by one level for publication bias because the funnel plot displayed asymmetry.
We graded the certainty of evidence as very low for the overall rate of surgical complications. We downgraded the certainty of evidence by one level for serious risk of bias because five studies had high risk of bias for blinding of participants and personnel (Bai 2016; Berger 2009; El Nakeeb 2015; Senda 2018; Singh 2018), and two studies had high risk of bias for blinding of outcome assessment (Berger 2009; Singh 2018). We downgraded the certainty of evidence by one level for serious unexplained inconsistency (moderate heterogeneity I2 = 45%). We downgraded the certainty of evidence by one level for serious imprecision because the confidence interval included both potential benefit and potential harm from the intervention. We downgraded the certainty of the evidence by one level for indirectness because of the different numbers of performing surgeons with different levels of experience and different proportions of participants who received pancreatic duct stents and somatostatin or its analogues among the included studies. We downgraded the certainty of evidence by one level for publication bias because the funnel plot displayed asymmetry.
We graded the certainty of evidence as very low for the length of hospital stay. We downgraded the certainty of evidence by two levels for very serious risk of bias because all included studies were at high risk of bias for blinding of participants and personnel and this outcome was determined largely by the performing surgeons. We downgraded the certainty of evidence by one level for serious inconsistency in the direction of effects across the studies. We downgraded the certainty of evidence by one level for serious imprecision because the wide confidence interval included both potential benefit and potential harm from the intervention. We downgraded the certainty of the evidence by one level for indirectness because of the different numbers of performing surgeons with different levels of experience and different proportions of participants who received pancreatic duct stents and somatostatin or its analogues among the included studies. We downgraded the certainty of evidence by one level for publication bias because the funnel plot displayed asymmetry.
For the comparison of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, we rated the certainty of evidence for six outcomes, including rate of POPF (grade B or C defined according to the 2016 ISGPS definition), postoperative mortality, rate of surgical reintervention, rate of postoperative bleeding, overall rate of surgical complications, and length of hospital stay.
We graded the certainty of evidence as very low for the rate of POPF (grade B or C defined according to the 2016 ISGPS definition), rate of surgical reintervention, rate of postoperative bleeding, overall rate of surgical complications, and length of hospital stay. We downgraded the certainty of evidence by two levels for very serious risk of bias because the study that evaluated this was at high risk of bias for blinding of participants and personnel, blinding of outcome assessment, and incomplete outcome data (Hirono 2019). We downgraded the certainty of the evidence by two levels for very serious imprecision because of the small sample size and wide confidence interval that included both potential benefit and potential harm from the intervention.
We graded the certainty of evidence as very low for postoperative mortality. Due to the nature of the interventions, the performing surgeons could not be blinded. The study was at high risk of bias for blinding of participants and outcome assessors; however, because unblinding of participants, performing surgeons, or outcome assessors is not expected to impact on postoperative mortality, we did not downgrade the certainty of evidence for unblinding of participants, performing surgeons, or outcome assessors. We downgraded the certainty of the evidence by one level for serious risk of bias because the study was at high risk of bias for incomplete outcome data. We downgraded the certainty of the evidence by two levels for very serious imprecision because of the small sample size and the wide confidence interval that included both potential benefit and potential harm from the intervention.
Potential biases in the review process
There were several potential biases of note in the review process. First, we performed a thorough and systematic literature search, but the possibility of missing relevant trial data remains because of potential publication bias. Second, some potentially relevant data were not reported in the included studies. We were able to obtain some of these data by contacting the authors of the studies, but despite repeated contact, some data remained missing. Third, the lack of a unified standard for the definition of POPF, in particular in the earlier studies (Bassi 2003; Berger 2009; Chou 1996), might introduce bias in this review. In order to minimize this bias, we analyzed only the outcome data of studies that used the 2005 ISGPF definition or the 2016 ISGPS definition of POPF. In addition, we observed no change in the rate of POPF in the sensitivity analysis by including studies that did or did not use the ISGPF definition of POPF.
Agreements and disagreements with other studies or reviews
Several systematic reviews have compared duct‐to‐mucosa pancreaticojejunostomy with invagination pancreaticojejunostomy. Five systematic reviews found that duct‐to‐mucosa pancreaticojejunostomy and invagination pancreaticojejunostomy were comparable in terms of the rate of POPF (grade B or C) (Bai 2013; Lyu 2018; Kilambi 2018; Sun 2016; Zhang 2017). The conclusions of these five reviews may not be justified because the CIs for the rate of POPF were very wide (Table 2). Two systematic reviews found that duct‐to‐mucosa pancreaticojejunostomy appeared to increase the rate of POPF (grade B or C) (Cao 2020; Hua 2015). However, none of the previous systematic reviews included Xu 2015, which favored duct‐to‐mucosa pancreaticojejunostomy. The conclusions presented in those two reviews (Cao 2020; Hua 2015), may have been changed by including Xu 2015. All previous systematic reviews included only studies that used either the 2005 ISGPF definition of POPF (Bai 2013; Hua 2015; Lyu 2018; Kilambi 2018; Sun 2016; Zhang 2017), or the 2016 ISGPS definition of POPF (Cao 2020), when evaluating the rate of POPF (Table 2).
| Review | Included studies (participants) | POPF definition | Rate of POPF (95% CI) | Postoperative mortality (95% CI) | Rate of surgical reintervention(95% CI) | Rate of postoperative bleeding (95% CI) | Overall rate of surgical complications (95% CI) | Length of hospital stay (95% CI) | GRADE assessment | Conclusions |
|---|---|---|---|---|---|---|---|---|---|---|
| 4 (467) (Berger 2009; Chou 1996; Han 2009; Langrehr 2005) | 2005 ISGPF | RR 0.93 (0.17 to 5.26)
| RR 1.18 (0.39 to 3.54) | RR 1.09 (0.54 to 2.22) | ‐ | RR 0.91 (0.69 to 1.21) | MD ‐1.78 (‐4.60 to 1.04) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 7 (1110) (Bai 2016; Berger 2009; El Nakeeb 2015;Maemura 2015 Senda 2018; Singh 2018; Xu 2015) | 2016 ISGPS | OR 1.78 (1.18 to 2.67) | OR 0.79 (0.36 to 1.75) | ‐ | ‐ | ‐ | MD ‐0.33 (‐1.80 to 1.14) | No | Duct‐to‐mucosa pancreaticojejunostomy is worse than invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2016 ISGPS definition) | |
| 5 (654) (Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Langrehr 2005) | 2005 ISGPF | OR 2.94 (1.31 to 6.60) | OR 1.12 (0.44 to 2.83) | OR 1.42 (0.74 to 2.74) | ‐ | OR 1.10 (0.80 to 1.50) | MD ‐0.54 (‐2.58 to 1.50) | No | Duct‐to‐mucosa pancreaticojejunostomy is worse than invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 8 (1099) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Langrehr 2005; Senda 2018; Singh 2018) | 2005 ISGPF | RR 1.14 (0.65 to 3.04) | RR 1.15 (0.67 to 1.97) | RR 0.94 (0.47 to 1.89) | RR 0.94 (0.44 to 2.00) | RR 1.06 (0.95 to 1.19) | MD ‐0.17 (‐2.56 to 2.23) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 8 (1043) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Singh 2018) | 2005 ISGPF | OR 0.95 (0.30 to 2.96) | OR 0.74 (0.36 to 1.53) | OR 1.10 (0.59 to 2.05) | OR 0.81 (0.45 to 1.45) | OR 0.98 (0.72 to 1.33) | MD ‐1.35 (‐2.91 to 0.22) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF | |
| 7 (850) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005) | 2005 ISGPF | RR 0.78 (0.15 to 3.96) | RR 0.94 (0.40 to 2.18) | ‐ | ‐ | RR 0.98 (0.82 to 1.16) | MD ‐2.80 (‐5.08 to 0.52) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 7 (850) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005) | 2005 ISGPF | RR 0.78 (0.15 to 3.96) | RR 1.01 (0.45 to 2.24) | RR 1.16 (0.67 to 2.02) | ‐ | RR 0.98 (0.86 to 1.12) | MD ‐1.54 (‐3.49 to 0.41) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| This review | 10 (1472) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015) | 2005 ISGPF | RR 1.44 (0.64 to 3.26) | RR 0.77 (0.39 to 1.49) | RR 1.12 (0.65 to 1.95) | RR 0.85 (0.51 to 1.42) | RR 1.12 (0.92 to 1.36) | MD ‐0.41 (‐1.87 to 1.04) | Yes | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) |
RR: risk ratio; OR: odds ratio; MD: mean difference; POPF: Postoperative pancreatic fistula; ISGPF: International Study Group on Pancreatic Fistula; ISGPS: International Study Group on Pancreatic Surgery; ‐: not available
We disagree with the findings of previous systematic reviews (Bai 2013; Cao 2020; Hua 2015; Kilambi 2018; Lyu 2018; Sun 2016; Zhang 2017). We found that the impact of duct‐to‐mucosa pancreaticojejunostomy on the rate of POPF (grade B or C) was very uncertain for people undergoing pancreaticoduodenectomy because the certainty of the evidence was very low based on the use of GRADE assessments in this review.
The differences between the conclusions presented in previous systematic reviews and the conclusions presented in this review may in large part be due to the addition of one trial (Xu 2015), which favored duct‐to‐mucosa pancreaticojejunostomy. An additional difference is that none of the previous reviews used GRADE to assess the certainty of the evidence (Table 2).
The study by Bai and colleagues seemed to be an outlier for the outcome of POPF (grade B or C defined according to the 2005 ISGPF definition) (Bai 2016). A possible reason for the difference between this study and the other included studies was that all pancreaticojejunostomies were performed by a single surgeon in Bai 2016.
Study flow diagram.
Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgments about each risk of bias item for each included study.
Funnel plot of comparison: 1 Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy, outcome: 1.2 Postoperative mortality.
Funnel plot of comparison: 1 Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy, outcome: 1.3 Rate of surgical reintervention.
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 1: Rate of postoperative pancreatic fistula (grade B or C defined according to the 2005 ISGPF definition)
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 2: Postoperative mortality
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 3: Rate of surgical reintervention
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 4: Rate of postoperative bleeding
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 5: Overall rate of surgical complications
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 6: Length of hospital stay
| Length of hospital stay | ||||
| Study | Number in study | Comparison | Results | Comment |
|---|---|---|---|---|
| Bai 2016 | 132 (64 versus 68) | Duct‐to‐mucosa versus invagination pancreaticojejunostomy | Duct‐to‐mucosa versus invagination pancreaticojejunostomy: median (range): 13 (6‐42) versus 15 (6‐88), p value = 0.021 | Authors reported that the length of hospital stay was shorter in the duct‐to‐mucosa group |
| Bassi 2003 | 144 (72 versus 72) | Duct‐to‐mucosa versus invagination pancreaticojejunostomy | Duct‐to‐mucosa versus invagination pancreaticojejunostomy: median: 16 versus 17, without reporting the p value | Authors reported that there was difference in the length of hospital stay between groups |
| Berger 2009 | 197 (197 versus 100) | Duct‐to‐mucosa versus invagination pancreaticojejunostomy | Duct‐to‐mucosa versus invagination pancreaticojejunostomy: median (range): 7.0 (5‐96) versus 7.0 (5‐169), without reporting the p value | Authors reported that there was no difference in the length of hospital stay between groups |
| El Nakeeb 2015 | 107 (53 versus 54) | Duct‐to‐mucosa versus invagination pancreaticojejunostomy | Duct‐to‐mucosa versus invagination pancreaticojejunostomy: median (range): 8.0 (5‐41) versus 8.0 (4‐35), p value = 0.83 | Authors reported that there was no difference in the length of hospital stay between groups |
| Langrehr 2005 | 113 (56 versus 57) | Duct‐to‐mucosa versus invagination pancreaticojejunostomy | Duct‐to‐mucosa versus invagination pancreaticojejunostomy: median: 15 versus 16, without reporting the p value | Authors reported that there was no difference in the length of hospital stay between groups |
| Senda 2018 | 120 (61 versus 59) | Duct‐to‐mucosa versus invagination pancreaticojejunostomy | Duct‐to‐mucosa versus invagination pancreaticojejunostomy: median (range): 24 (9‐107) versus 19 (9‐77), p value = 0.015 | Authors reported that the length of hospital stay was shorter in the invagination group |
Comparison 1: Duct‐to‐mucosa versus invagination pancreaticojejunostomy, Outcome 7: Length of hospital stay
Comparison 2: Duct‐to‐mucosa versus invagination pancreaticojejunostomy (sensitivity analysis including studies that did or did not use the ISGPF definition of POPF), Outcome 1: Rate of postoperative pancreatic fistula
Comparison 3: Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, Outcome 1: Rate of postoperative pancreatic fistula (grade B or C defined according to the 2016 ISGPS definition)
Comparison 3: Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, Outcome 2: Postoperative mortality
Comparison 3: Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, Outcome 3: Rate of surgical reintervention
Comparison 3: Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, Outcome 4: Rate of postoperative bleeding
Comparison 3: Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, Outcome 5: Overall rate of surgical complications
| Length of hospital stay | ||||
| Study | Number in study | Comparison | Results | Comment |
|---|---|---|---|---|
| Hirono 2019 | 210 (107 versus 103) | Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus the interrupted suture technique | Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus the interrupted suture technique: median (range): 15 (8‐52) versus 15 (6‐44), p value = 0.104 | Authors reported that there was no difference in the length of hospital stay between groups |
Comparison 3: Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique, Outcome 6: Length of hospital stay
Comparison 4: Duct‐to‐mucosa versus invagination pancreaticojejunostomy (stratified by pancreatic texture), Outcome 1: Rate of postoperative pancreatic fistula (grade A or B or C defined according to the 2005 ISGPF definition)
Comparison 4: Duct‐to‐mucosa versus invagination pancreaticojejunostomy (stratified by pancreatic texture), Outcome 2: Postoperative mortality
Comparison 5: Duct‐to‐mucosa versus invagination pancreaticojejunostomy (sensitivity analysis using the fixed‐effect model), Outcome 1: Rate of postoperative pancreatic fistula (grade B or C defined according to the 2005 ISGPF definition)
| Duct‐to‐mucosa compared to invagination pancreaticojejunostomy for the prevention of postoperative pancreatic fistula following pancreaticoduodenectomy | ||||||
| Patient or population: adults undergoing open pancreaticoduodenectomy | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with invagination pancreaticojejunostomy | Risk with duct‐to‐mucosa pancreaticojejunostomy | |||||
| Rate of POPF (grade B or C defined according to the 2005 ISGPF definition) Follow‐up: 30 days | 80 per 1000 | 115 per 1000 | RR 1.45 | 1122 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of POPF (grade B or C defined according to the 2005 ISGPF definition). |
| Postoperative mortality Follow‐up: 30 days | 26 per 1000 | 20 per 1000 | RR 0.77 | 1472 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on postoperative mortality. |
| Adverse events | This outcome was not reported in any of the included studies. | |||||
| Rate of surgical reintervention Follow‐up: 30 days | 34 per 1000 | 38 per 1000 | RR 1.12 | 1472 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of surgical reintervention. |
| Rate of postoperative bleeding Follow‐up: 30 days | 47 per 1000 | 40 per 1000 | RR 0.85 | 1275 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the rate of postoperative bleeding. |
| Overall rate of surgical complications Follow‐up: 30 days | 484 per 1000 | 542 per 1000 | RR 1.12 | 750 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the overall rate of surgical complications. |
| Length of hospital stay (days) | The mean length of hospital stay was 16 to 22 days | The mean length of hospital stay was 14.9 to 20 days | MD ‐0.41 (‐1.87 to 1.04) | 658 | ⊕⊝⊝⊝ | Six additional studies with skewed data were not suitable for pooling. One trial reported that the duct‐to‐mucosa group was associated with less time in hospital, but another trial reported that the duct‐to‐mucosa group was associated with more time in hospital than the invagination group. The other four studies reported no difference in the length of hospital stay between groups. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy on the length of hospital stay. |
| Quality of life | This outcome was not reported in any of the included studies. | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded one level for serious risk of bias: all included studies with high risk of bias for blinding of participants and personnel and some of the studies had high risk of bias for blinding of outcome assessment. | ||||||
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy usingthe traditional interrupted technique for the prevention of postoperative pancreatic fistula following pancreaticoduodenectomy | ||||||
| Patient or population: adults undergoing open pancreaticoduodenectomy | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Risk with duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique | |||||
| Rate of POPF (grade B or C defined according to the 2016 ISGPS definition) Follow‐up: 30 days | 68 per 1000 | 103 per 1000 | RR1.51 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of POPF (grade B or C defined according to the 2016 ISGPS definition) compared with duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique. |
| Postoperative mortality Follow‐up: 30 days | See comment | See comment | Not estimable | 210 | ⊕⊝⊝⊝ | There were no postoperative deaths in either group. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on postoperative mortality. |
| Adverse events | This outcome was not reported in the study. | |||||
| Rate of surgical reintervention Follow‐up: 30 days | 10 per 1000 | 19 per 1000 | RR 1.93 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of surgical reintervention. |
| Rate of postoperative bleeding Follow‐up: 30 days | ‐ | ‐ | RR 2.89 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the rate of postoperative bleeding. |
| Overall rate of surgical complications Follow‐up: 30 days | 408 per 1000 | 449 per 1000 | RR 1.10 | 210 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the overall rate of surgical complications . |
| Length of hospital stay (days) | The median length of hospital stay was 15 days | The median length of hospital stay was 15 days | See comment | 210 | ⊕⊝⊝⊝ | Authors reported that there was no difference in the length of hospital stay between groups. The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique compared to duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique on the length of hospital stay. |
| Quality of life | This outcome was not reported in the study. | |||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded two levels for very serious risk of bias: the study was at high risk of bias for blinding of participants and personnel, blinding of outcome assessment, and incomplete outcome data. | ||||||
| Comparisons | Outcomes | Main analysis (95% CI) | Sensitivity analysis using fixed‐effect model (95% CI) | Sensitivity analysis by changing between worst‐case scenario analysis and best‐case scenario analysis for missing data | Posthoc sensitivity analysis by including studies that did or did not use the ISGPF definition of POPF | |
|---|---|---|---|---|---|---|
| Worst/best‐case (95% CI) | Best/worst‐case (95% CI) | |||||
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Rate of POPF (grade B or C defined according to the 2005 ISGPF definition) | RR 1.44 (0.64 to 3.26) | RR 1.50 (1.05 to 2.15) | RR 1.44 (0.64 to 3.26) | RR 1.44 (0.64 to 3.26) | RR 1.13 (0.60 to 2.15) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Postoperative mortality | RR 0.77 (0.39 to 1.49) | RR 0.81 (0.44 to 1.52) | RR 0.77 (0.39 to 1.49) | RR 0.77 (0.39 to 1.49) | RR 0.77 (0.39 to 1.49) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Adverse events | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Rate of surgical reintervention | RR 1.12 (0.65 to 1.95) | RR 1.01 (0.61 to 1.70) | RR 1.12 (0.65 to 1.95) | RR 1.12 (0.65 to 1.95) | RR 1.12 (0.65 to 1.95) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Rate of postoperative bleeding | RR 0.85 (0.51 to 1.42) | RR 0.87 (0.53 to 1.44) | RR 0.85 (0.51 to 1.42) | RR 0.85 (0.51 to 1.42) | RR 0.85 (0.51 to 1.42) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Overall rate of surgical complications | RR 1.12 (0.92 to 1.36) | RR 1.11 (0.97 to 1.28) | RR 1.12 (0.93 to 1.36) | RR 1.12 (0.93 to 1.36) | RR 1.12 (0.92 to 1.36) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Length of hospital stay | MD ‐0.41 (‐1.87 to 1.04) | MD ‐0.41 (‐1.87 to 1.04) | ‐ | ‐ | MD ‐0.41 (‐1.87 to 1.04) |
| Duct‐to‐mucosa pancreaticojejunostomy versus invagination pancreaticojejunostomy | Quality of life | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Rate of POPF (grade B or C defined according to the 2016 ISGPS definition) | RR 1.51 (0.61 to 3.75) | RR 1.51 (0.61 to 3.75) | RR 2.29 (0.98 to 5.34) | RR 0.69 (0.33 to 1.41) | RR 1.51 (0.61 to 3.75) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Postoperative mortality | No events | No events | RR 11.00 (0.62 to 196.60) | RR 0.05 (0.00 to 0.89) | No events |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Adverse events | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Rate of surgical reintervention | RR 1.93 (0.18 to 20.91) | RR 1.93 (0.18 to 20.91) | RR 7.00 (0.88 to 55.97) | RR 0.20 (0.04 to 0.89) | RR 1.93 (0.18 to 20.91) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Rate of postoperative bleeding | RR 2.89 (0.12 to 70.11) | RR 2.89 (0.12 to 70.11) | RR 13.00 (0.74 to 228.05) | RR 0.11 (0.01 to 0.86) | RR 2.89 (0.12 to 70.11) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Overall rate of surgical complications | RR 1.10 (0.80 to 1.51) | RR 1.10 (0.80 to 1.51) | RR 1.26 (0.93 to 1.72) | RR 0.94 (0.70 to 1.26) | RR 1.10 (0.80 to 1.51) |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Length of hospital stay | ‐ | ‐ | ‐ | ‐ | ‐ |
| Duct‐to‐mucosa pancreaticojejunostomy using the modified Blumgart technique versus duct‐to‐mucosa pancreaticojejunostomy using the traditional interrupted technique | Quality of life | ‐ | ‐ | ‐ | ‐ | ‐ |
| CI: confidence interval ; RR: risk ratio; MD: mean difference; POPF: postoperative pancreatic fistula; ISGPF: International Study Group on Pancreatic Fistula; ISGPS: International Study Group on Pancreatic Surgery; ‐: not available | ||||||
| Review | Included studies (participants) | POPF definition | Rate of POPF (95% CI) | Postoperative mortality (95% CI) | Rate of surgical reintervention(95% CI) | Rate of postoperative bleeding (95% CI) | Overall rate of surgical complications (95% CI) | Length of hospital stay (95% CI) | GRADE assessment | Conclusions |
|---|---|---|---|---|---|---|---|---|---|---|
| 4 (467) (Berger 2009; Chou 1996; Han 2009; Langrehr 2005) | 2005 ISGPF | RR 0.93 (0.17 to 5.26)
| RR 1.18 (0.39 to 3.54) | RR 1.09 (0.54 to 2.22) | ‐ | RR 0.91 (0.69 to 1.21) | MD ‐1.78 (‐4.60 to 1.04) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 7 (1110) (Bai 2016; Berger 2009; El Nakeeb 2015;Maemura 2015 Senda 2018; Singh 2018; Xu 2015) | 2016 ISGPS | OR 1.78 (1.18 to 2.67) | OR 0.79 (0.36 to 1.75) | ‐ | ‐ | ‐ | MD ‐0.33 (‐1.80 to 1.14) | No | Duct‐to‐mucosa pancreaticojejunostomy is worse than invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2016 ISGPS definition) | |
| 5 (654) (Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Langrehr 2005) | 2005 ISGPF | OR 2.94 (1.31 to 6.60) | OR 1.12 (0.44 to 2.83) | OR 1.42 (0.74 to 2.74) | ‐ | OR 1.10 (0.80 to 1.50) | MD ‐0.54 (‐2.58 to 1.50) | No | Duct‐to‐mucosa pancreaticojejunostomy is worse than invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 8 (1099) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Langrehr 2005; Senda 2018; Singh 2018) | 2005 ISGPF | RR 1.14 (0.65 to 3.04) | RR 1.15 (0.67 to 1.97) | RR 0.94 (0.47 to 1.89) | RR 0.94 (0.44 to 2.00) | RR 1.06 (0.95 to 1.19) | MD ‐0.17 (‐2.56 to 2.23) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 8 (1043) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Singh 2018) | 2005 ISGPF | OR 0.95 (0.30 to 2.96) | OR 0.74 (0.36 to 1.53) | OR 1.10 (0.59 to 2.05) | OR 0.81 (0.45 to 1.45) | OR 0.98 (0.72 to 1.33) | MD ‐1.35 (‐2.91 to 0.22) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF | |
| 7 (850) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005) | 2005 ISGPF | RR 0.78 (0.15 to 3.96) | RR 0.94 (0.40 to 2.18) | ‐ | ‐ | RR 0.98 (0.82 to 1.16) | MD ‐2.80 (‐5.08 to 0.52) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| 7 (850) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005) | 2005 ISGPF | RR 0.78 (0.15 to 3.96) | RR 1.01 (0.45 to 2.24) | RR 1.16 (0.67 to 2.02) | ‐ | RR 0.98 (0.86 to 1.12) | MD ‐1.54 (‐3.49 to 0.41) | No | Duct‐to‐mucosa pancreaticojejunostomy is comparable to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) | |
| This review | 10 (1472) (Bai 2016; Bassi 2003; Berger 2009; Chou 1996; El Nakeeb 2015; Han 2009; Langrehr 2005; Senda 2018; Singh 2018; Xu 2015) | 2005 ISGPF | RR 1.44 (0.64 to 3.26) | RR 0.77 (0.39 to 1.49) | RR 1.12 (0.65 to 1.95) | RR 0.85 (0.51 to 1.42) | RR 1.12 (0.92 to 1.36) | MD ‐0.41 (‐1.87 to 1.04) | Yes | The evidence is very uncertain about the effect of duct‐to‐mucosa pancreaticojejunostomy compared to invagination pancreaticojejunostomy in terms of the rate of POPF (grade B or C using the 2005 ISGPF definition) |
| RR: risk ratio; OR: odds ratio; MD: mean difference; POPF: Postoperative pancreatic fistula; ISGPF: International Study Group on Pancreatic Fistula; ISGPS: International Study Group on Pancreatic Surgery; ‐: not available | ||||||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Rate of postoperative pancreatic fistula (grade B or C defined according to the 2005 ISGPF definition) Show forest plot | 7 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 1.1.1 Grade B or C | 7 | 1122 | Risk Ratio (M‐H, Random, 95% CI) | 1.45 [0.64, 3.26] |
| 1.1.2 Grade B | 7 | 1122 | Risk Ratio (M‐H, Random, 95% CI) | 1.41 [0.63, 3.15] |
| 1.1.3 Grade C | 7 | 1122 | Risk Ratio (M‐H, Random, 95% CI) | 1.45 [0.58, 3.61] |
| 1.2 Postoperative mortality Show forest plot | 10 | 1472 | Risk Ratio (M‐H, Random, 95% CI) | 0.77 [0.39, 1.49] |
| 1.3 Rate of surgical reintervention Show forest plot | 10 | 1472 | Risk Ratio (M‐H, Random, 95% CI) | 1.12 [0.65, 1.95] |
| 1.4 Rate of postoperative bleeding Show forest plot | 9 | 1275 | Risk Ratio (M‐H, Random, 95% CI) | 0.85 [0.51, 1.42] |
| 1.5 Overall rate of surgical complications Show forest plot | 5 | 750 | Risk Ratio (M‐H, Random, 95% CI) | 1.12 [0.92, 1.36] |
| 1.6 Length of hospital stay Show forest plot | 4 | 658 | Mean Difference (IV, Random, 95% CI) | ‐0.41 [‐1.87, 1.04] |
| 1.7 Length of hospital stay Show forest plot | 6 | Other data | No numeric data | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Rate of postoperative pancreatic fistula Show forest plot | 10 | 1472 | Risk Ratio (M‐H, Random, 95% CI) | 1.13 [0.60, 2.15] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 3.1 Rate of postoperative pancreatic fistula (grade B or C defined according to the 2016 ISGPS definition) Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 3.1.1 Grade B or C | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | 1.51 [0.61, 3.75] |
| 3.1.2 Grade B | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | 1.51 [0.61, 3.75] |
| 3.1.3 Grade C | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | Not estimable |
| 3.2 Postoperative mortality Show forest plot | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | Not estimable |
| 3.3 Rate of surgical reintervention Show forest plot | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | 1.93 [0.18, 20.91] |
| 3.4 Rate of postoperative bleeding Show forest plot | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | 2.89 [0.12, 70.11] |
| 3.5 Overall rate of surgical complications Show forest plot | 1 | 210 | Risk Ratio (M‐H, Random, 95% CI) | 1.10 [0.80, 1.51] |
| 3.6 Length of hospital stay Show forest plot | 1 | Other data | No numeric data | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 4.1 Rate of postoperative pancreatic fistula (grade A or B or C defined according to the 2005 ISGPF definition) Show forest plot | 5 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 4.1.1 Soft pancreas | 5 | 505 | Risk Ratio (M‐H, Random, 95% CI) | 1.64 [0.83, 3.24] |
| 4.1.2 Firm pancreas | 4 | 316 | Risk Ratio (M‐H, Random, 95% CI) | 1.08 [0.54, 2.18] |
| 4.2 Postoperative mortality Show forest plot | 2 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 4.2.1 Soft pancreas | 2 | 208 | Risk Ratio (M‐H, Random, 95% CI) | 1.59 [0.20, 12.64] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 5.1 Rate of postoperative pancreatic fistula (grade B or C defined according to the 2005 ISGPF definition) Show forest plot | 7 | 1121 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.50 [1.05, 2.14] |
