Conservative approach in the management of severe acute pancreatitis: eight-year experience in a single institution

Abstract

Introduction. Recognition of severe acute pancreatitis (SAP), intensive care, shifting away from early surgical treatment, and monitoring of the intra-abdominal pressure (IAP) is important in the management of SAP. The aim of our study was retrospective evaluation and critical assessment of the experience with SAP management protocol involving new strategy in the university hospital. Methods. Protocols of 274 SAP patients treated in our institution during the last eight years were reassessed. APACHE II, CRP and SOFA score, IAP, pulmonary complications, ventilatory support and infection rate were evaluated. The success of the conservative treatment, surgical interventions and mortality was analysed comparing period 1 from 1999 to 2002 and period 2 from 2003 to 2006. Results. More patients with necrotising SAP were treated in period 2. The average CRP and SOFA score was higher in period 2, p=0.018; p=0.011. A total of 139 patients underwent continuous veno-venouse haemofiltration (CVVH) as a component of fluid resuscitation and IAP control. Application of CVVH increased in period 2, p<0.005. Only 5–8% of patients were managed with ventilatory support. The overall infection rate decreased in period 2 comprising 21%, p<0.005. Success rate of the conservative therapy reached 69% in period 2, p<0.01. Surgical treatment was performed in 41% of patients in period 1 vs. 19% in period 2, p<0.001. Overall mortality was 19%, with a reduction to 12% in year 2006. Conclusion. The conservative protocol-based approach is a rational treatment strategy for the management of SAP and can be successfully implemented in the setting of the university hospital.

Introduction

Management of acute pancreatitis is a hot topic through the recent decades. The basic treatment strategy is highlighted in several international publications and generally recommends prompt recognition of severe forms, aggressive intensive care and a shift away from early surgical treatment in necrotising forms ¹. A considerable part of patients including those with necrotising forms can be successfully treated conservatively ². A more conservative approach gives opportunity to postpone surgical intervention in necrotising forms in the major part of patients on average after the third week from the onset of the disease in cases when infection and sepsis are not clinically manifest ³. Formation of postnecrotic pseudocysts is one possible form of outcome after in-hospital treatment. Involvement of interventional therapies during the early treatment course is questionable due to associated risk of nosocomial infection; however, it could be rational when applied for the initial control of sepsis for patients in poor condition. This approach is associated with less dramatic systemic response; at the same time, the clinical course can later be associated with progression of systemic inflammatory response syndrome (SIRS) and MODS mainly as a consequence of incomplete drainage of the solid necrotic infected tissue ¹,⁴. The aim of our study was retrospective evaluation and critical assessment of the experience with implementation of severe acute pancreatitis (SAP) management protocol involving new strategy and treatment modalities available in our university hospital.

Methods

Protocols of patients who suffered SAP and were treated in our institution during the time period between year 1999 and 2006 were reassessed considering the implementation of more conservative approach, early application of continuous veno-venouse haemofiltration (CVVH), early enteral feeding, and control of the intra-abdominal hypertension (IAH). Patients with malignancies were not included.

Diagnosis and severity assessment

Acute pancreatitis was diagnosed according to the clinical presentation of the disease, threefold elevation of the lipase in blood and typical signs on visual diagnostics. Ultrasound scans for detection of gallstone disease, fluid collections, plain chest and abdominal X-ray were routinely used at admission and during the treatment course. Initial severity assessment was done after clinical evaluation, APACHE II scoring, evidence of SIRS and/or organ dysfunction at admission. SIRS was diagnosed if two or more of the following conditions were detected: (a) temperature above 38°C or below 36°C; (b) heart rate above 90 beats/min; (c) respiratory rate above 20 breath/min; or arterial carbon dioxide tension below 32 mmHg; and (d) white blood cell count above 12,000/mm³ or below 4000/mm³, or immature (band) forms accounting for more than 10% of neutrophils present ⁵. The intensive care specialist was asked to assess patients with initially suspected SAP for transfer to the ICU. SAP was classified according to the Atlanta 1992 criteria ⁶. Severe manifestation of the disease was considered when the clinical picture, threefold elevation of serum lipase accompanied by persistent SIRS and/or organ dysfunction, and CRP > 150 mg/day during the first 48 hours of initial treatment were recognised. Organ dysfunction was defined according to the recommendations of the Consensus Conference of American College of Chest Physicians/Society of Critical Care medicine in 1991 ⁷,⁸. A MODS was diagnosed if dysfunction of more than one organ was detected, requiring intervention to maintain homeostasis. SOFA score was calculated for daily assessment of the dynamics of the organ dysfunction ⁸. Intra-abdominal pressure (IAP) measurement was done indirectly through the urinary bladder ⁹ from year 2000 as part of the ICU treatment protocol. IAH was considered when IAP exceeded 12 mmHg according to the World Society Abdominal Compartment Syndrome definition ¹⁰.

Diagnosis of necrotising SAP was based on the clinical picture supported by the radiological evidence of pancreatic necrosis and/or involvement of the peripancreatic tissue on the contrast-enhanced CT scan considering the period of at least 96 hours after the onset of the disease. First contrast-enhanced CT scan was postponed until the patient was transferred from ICU to the surgical department when the clinical diagnosis of SAP was evident and conservative treatment was effective. Repeated contrast-enhanced CT scan was performed in cases of suspected infection or clinical deterioration. Early contrast-enhanced CT scan was performed for differential diagnosis with other acute surgical pathology. In some cases diagnosis was proved during the operation when early surgical intervention was performed. The presence of infection was suspected when signs of sepsis and clinical deterioration with negative dynamics of the organ dysfunction were evident and proved by contrast-enhanced CT scan. Incidence of septic complications was analysed considering the possible way of contamination of the necrotic tissue and/or peripancreatic fluid. Two types of peripancreatic infections were defined – infection of necrotic tissue and peripancreatic fluid collections prior to surgery and infection after early surgery. Sepsis with positive blood cultures and patients who demonstrated signs of infection being admitted within the second phase of the disease were analysed separately.

Procalcitonin test was implemented since year 2005. Fine needle aspiration and bacterial culture of the aspirate were not used on routine basis.

Success rate of the conservative or the surgical treatment was compared in two periods – period 1 from 1999 to 2002 and period 2 from 2003 to 2006. Frequency of the ventilatory support, pneumonia, pleural exudate, pulmonary and/or renal dysfunction, IAH, and infection rate were recorded in both periods. The number of surgical interventions, overall ICU, hospital stay, and mortality was analysed.

Treatment protocol

SAP treatment protocol was accepted by hospital administration and the local ethics commission, and patients were informed about the procedures and the associated risks and complications. Initial treatment consisted of fluid replacement including colloids to achieve isovolemic haemodilution and organ support ¹¹. All patients who were treated only by means of conservative therapy received early oral feeding (EOF) with whole protein enteral formula as part of treatment protocol when gastroenteric transit was not severely impaired. Initial oral feeding was started with 20 ml bolus of enteral formula every two hours as a test feeding in patients who explicitly accepted the feeding protocol and were able to cooperate in the treatment procedure. The feeding rate was advanced according to individual tolerance by 20 ml bolus feeds every hour with further increase of the feeding volume. The initially employed feed was nutritionally complete, low fat, whole protein formulas containing 1 kcal/ml (Nutrison Standard or Nutricomp). We did not follow the principle of the “goal feeding” but rather the principle “to feed the gut”. In more complicated cases jejunal or gastric tube feeding was applied and complemented with parenteral nutrition and specially adapted formulas when indicated. Parenteral nutrition was not used when enteral or oral feeding was commenced in 3–5 days and the clinical course improved.

The main indications for application of CVVH were progression of SIRS, abdominal compartment syndrome (IAP > 20 mmHg) and persistence of MODS despite complex initial 24–48 hour intensive therapy.

CVVH parameters – haemofiltration (HF) was performed using Diapact CRRT from B.Braun Co or Fresenius Medical Care Multifiltrate machines, Germany. The substitution fluid was infused at the rate of 1000 ml/h in pre-diluted or post-diluted manner, comprising 24–35 l of total substitute in 24 hours. The blood flow ranged within 50–200 ml/min. Ultra filtration rate was controlled depending on diuresis and fluid balance. Haemodiafiltration was provided in accordance with individual indications. Antibiotic prophylaxis was used for patients with severely compromised tissue perfusion, early progressive MODS, sustained increase of the IAP and signs of cholangitis. The indication for early surgical intervention was mainly deterioration of the clinical course despite initial conservative treatment, progression of ACS and obscure diagnosis on admission. Infection was the main indication for surgery after the second week and later.

Statistics

Statistical comparison was performed with paired samples T-test. Clinical data were expressed as average±standard deviation. Correlation data were calculated using Pearson's and Spearman's Correlation test. Data analysis was performed with SPSS software version 11.0 (SPSS Inc.).

Results

A total of 274 patients were included in our retrospective review, with their age ranging from 19 to 84 years. Male to female ratio was 2.8:1. The main aetiologic factor was alcohol in 54%, predominantly in male patients, and gallstone disease in 19%, predominantly in female patients. Four patients developed SAP in association with abdominal trauma. Significant hyperlipidemia was observed in some patients; however, it was related to fat meal and/or alcohol consumption in most cases. Average APACHE II score on admission was 7.4±4.29 and 8.2±5.57 in period 1 and period 2, respectively. One hundred and thirty-two patients suffered SAP without evident necrosis. Necrotising SAP was recognised in 142 patients with increased incidence during the years 2003–2006. Systemic inflammatory response as a pathophysiologic consequence of SAP developed with similar incidence through both periods in 93–96% of patients. CRP – the biochemical marker of SIRS had typical dynamics in both periods with the average peak value on day 3 after admission and decline on day 7; however, significantly higher average CRP was observed on admission in period 2, p=0.018 (Figure 1 and Table II). Impairment of the organ function with involvement of at least two organ systems defined as MODS complicated the clinical course in 85–90% of patients (Table I). Dynamics of the organ dysfunction according to SOFA score resembled that of CRP; however, significantly higher average values were observed on admission, p=0.011, and on day 3, p=0.008, in period 2 (Figure 2 and Table II). Majority of patients were treated in ICU, though it was accepted as a clinical routine only in period 2. About half of the patients received ICU treatment directly from the day of admission more frequently in period 2 (Table I).

Figure 1. .

Dynamics of CRP.

Table II. CRP and SOFA score.

	Admission	Day 3	Day 7	Day14	Discharge
Average CRP, mg/l
1999–2002	102±83.68	185±110.21	96±68.85	67±63.10	42±51.35
2003–2006	183±146.44	263±122.87	141±63.08	88±67.70	48±53.01
P	0.018	NS	NS	NS	NS
SOFA score, points
1999–2002	2.5±1,59	2.1±1.61	0.8±1.19	0.3±0.46	0
2003–2006	3.6±2.62	3.7±2.61	1.4±2.00	1.2±1.75	0
P	0.011	0.008	NS	NS	NS

Table I. Epidemiology and severity of the disease.

	Total	1999–2002	2003–2006
SAP, No. of patients	274	160	114
Mean age (years)	47±14.92	49±15.76	44±13.12
Age range (years)	19–84	19–84	19–79
Male/female, No. /ratio		202/72/2.8: 1
Aetiology, No. of patients (%)
Alcohol	148 (54%)
Gallstones	52 (19%)
Other	74 (27%)
Time from onset (average days/hours)	2.5±2.43/65.0±61.18
Edematous SAP, No. of patients (%)	132	84 (53%)	48 (42%)
Necrotising SAP, No. of patients (%)	142	76 (48%)	66 (58%)
APACHE II score on admission		7.4±4.29	8.2±5.57
SIRS, No. of cases (%)		148 (93%)	109 (96%)
MODS, No. of cases (%)		136 (85%)	103 (90%)
ICU treatment		79%	100%
ICU treatment from admission		48%	56%

Figure 2. .

Dynamics of SOFA.

Antibiotic prophylaxis was provided when severe disturbances of the visceral perfusion were evident, and when multiple localisations of the inflammatory fluid collections, signs of early pulmonary, renal dysfunction, and significant elevation of IAP were diagnosed during the first 24 hours from hospitalisation. Ciprofloxacin combination with Metronidasol or Imipenem/Cilastin was used as an initial antibacterial therapy. Patients with signs of sepsis and cholangitis received antibacterial treatment routinely. The mode of early enteral feeding changed from predominantly jejunal tube feeding in the period 1 to early oral in period 2. Tube feeding including nasogastric way was used mainly in patients who needed longer nutritional support. Combined enteral and parenteral nutrition was considered in patients with severe catabolism and limited possibilities to provide adequate enteral nutrition.

Interventional therapy included drainage of the larger peripancreatic, retroperitoneal or free abdominal inflammatory exudate collections under ultrasound guidance for the temporary control of sepsis or IAH. Endoscopic sphincterotomy was performed in patients with progressive cholangitis and mechanical jaundice. Implementation of the SAP treatment protocol resulted in more routine early application of the CVVH in the period 2 comprising 87% vs. 25% in period 1, p<0.005 (Table III). Correlation analysis did not reveal interrelation between CVVH and infection therefore proving safety of the procedure.

Table III. CVVH and IAH.

	1999–2002	2003–2006	P
CVVH procedure (No. of patients,%)	40 (25%)	99 (87%)	<0.005
IAP measurements (No. of patients,%)	49 (31%)	16 (14%)	<0.01
IAH Grade I–II (No. of patients,%)	30 (19%)	11 (10%)	NS
IAH Grade III–IV (No. of patients,%)	19 (12%)	5 (4%)	0.028
ACS (No. of patients,%)	18 (11%)	5 (4%)	0.028
Correlation analysis, overall period	R	P
IAH: MODS	+0.188	0.002
IAH: Atelectasis	+0.154	0.001
IAH: Pulmonary dysfunction	+0.178	0.003
IAH: Pneumonia	+0.143	0.018
IAH Grade III–IV: Conservative treatment	−0.196	0.001
IAH Grade III–IV: Surgical treatment	+0.196	0.001

More cases of IAH were detected in period 1, p<0.01 including IAP corresponding to Grade I–II with significant difference for Grade III–IV, p<0.028 and ACS, p<0.028 (Table III). Positive interrelation was observed between IAH and incidence of MODS, r= + 0.188, p<0.002. IAH Grade III–IV positively interrelated with the frequency of surgical treatment (r= + 0.196, p<0.001) and negatively interrelated with application of conservative treatment only (r= − 0.196, p<0.001).

Implementation of the SAP treatment protocol was associated with significant increase in the recognition of pulmonary complications. Incidence of the revealed pleural exudates and pneumonia increased in the second period, p<0.005; p<0.001. At the same time, the number of atelectasis and incidence of pulmonary dysfunction according to SOFA score criteria were not statistically different on admission or later comparing both periods. Ventilatory support was applied in 5% of patients during the first period and in 8% during period 2 (Table IV). Correlation analysis revealed that pleural effusion and pulmonary dysfunction positively interrelated with SIRS throughout the whole period, r= + 0.142, p<0.019; r= + 0.19, p<0.002. Pulmonary complications including atelectasis, pulmonary dysfunction and pneumonia were also in positive interrelation with IAP, r= + 0.154, p<0.001; r= + 0.178, p<0.003; r= + 0.143, p<0.018. Pleural effusion positively interrelated with application of CVVH, r= + 0.393, p<0.005.

Table IV. Complications.

	1999–2002	2003–2006	P
Ventilatory support, No. of patients (%)	8 (5%)	9 (8%)
Pulmonary dysfunction, No. of patients (%)	72 (45%)	41 (36%)	NS
On admission	15 (9%)	10 (9%)	NS
Later	57 (36%)	31 (27%)	NS
Atelectasis, No. of patients (%)	10 (6%)	6 (5%)	NS
Pleural effusion, No. of patients (%)	41 (26%)	65 (57%)	<0.005
Pneumonia, No. of patients (%)	16 (10%)	32 (28%)	<0.001
Renal dysfunction, No. of patients (%)	45 (28%)	37 (33%)	NS
On admission	11 (7%)	10 (9%)	NS
Later	34 (21%)	27 (24%)	NS
Infection
Infection, No. of patients (%)	66 (41%)	24 (21%)	<0.005
CNIT Infection, No. of patients (%)	5 (3%)	12 (11%)	0.01
SIT Infection No. of patients (%)	61 (38%)	12 (11%)	<0.005
Admitted with infection	Four patients

Renal dysfunction was observed equally frequently in both periods at the time of admission and later during the treatment course despite the fact that more necrotising forms characterised period 2 (Table IV).

The Protocol treatment had an impact on the infection structure. Incidence of infection associated with drains and catheters due to early surgical or interventional therapy in the first period was significantly higher than cases of infection of necrotic tissue and peripancreatic fluidcollections prior to surgery comprising 38% vs. 3%, p<0.001.This disproportion changed in the second period when infection after early surgery decreased to the level of infection of necrotic tissue and peripancreatic fluid collections prior to surgery equally comprising 11%. The more conservative approach resulted in the decrease of the overall infection rate from 41% in period 1 vs. 21% in period 2, p<0.005 (Table IV) and accordingly in a decreased number of surgical interventions from 41 to 19%, p<0.001 (Table VI).

Table VI. Main outcomes.

	1999–2002	2003–2006	P
Hospital stay (days)	25.0±23.15	23.6±21.04	NS
ICU stay (days)	8.9±13.05	10.5±11.19	NS
Conservative treatment successful, No. of patients (%)	75 (47%)	79 (69%)	<0.01
Postnecrotic pseudocysts, No. of patients (%)	11 (7%)	15 (13%)	<0.05
Operations, No. of patients (%)	66 (41%)	22 (19%)	0.001
Mortality
Postoperative	21%	32%
Conservative treatment	20%	14%
Overall, No. of patients (%)	33 (21%)	20 (18%)	NS

Enterococci, Staphylococcus epidermidis, Klebsiella pneumoniae, Pseudomonas aeruginosa, Acinteobacter baumanii, Escherichia coli and Stenotrophonoma maltophilia were most often found in the positive cultures retrieved from the infected pancreatic and peripancreatic tissue. S. epidermidis, K. pneumoniae, Enterococci and E. coli were a more frequent finding in positive blood culture (Table V).

Table V. Bacterial flora.

Species	Positive cultures, No. of isolates (%)		Species	Positive cultures, No. of isolates (%)
Enterococcus	20	57	Staphylococcus hominis	1	3
Staphylococcus epidermidis	12	34	Moraxella lacunata	1	3
Klebsiella pneumoniae	7	20	Pseudomonas putida	1	3
Pseudomonas aeruginosa	6	17	Candida tropicans	1	3
Acinteobacter baumanii	5	14	Candida albicans	1	3
E. coli	4	11	Gemella morbillorum	1	3
Stenotrophomona maltophilia	4	11	MRSA	1	3
Staphylococcus aureus	3	9	Streptococcus oralis	1	3
Staphylococcus xylosus	2	6	Citrobacter	1	3
Enterobacter cloacae	2	6	Acinetobacter Iwoffii	1	3
Positive blood cultures
Staphylococcus epidermidis	9	64	Staphylococcus chromogenes	1	7
Klebsiella pneumoniae	2	14	Candida lipolytica	1	7
Enterococcus	2	14	Enterobacter cloacae	1	7
E. coli	2	14	Enterococcus cancerogenus	1	7
Criptococcus humicolus	1	7	MRSA	1	7
Acinteobacter baumanii	1	7	Staphylococcus aureus	1	7
Candida albicans	1	7

Successes of the conservative treatment significantly increased in period 2, when 69% of patients were discharged without any surgical intervention compared to 47% in period 1, p<0.01. As a consequence, formation of more postnecrotic pseudocysts was observed in the period 2, p<0.05. Although surgical treatment in period 2 was associated with relative but not significant increase in mortality, as most operations were done in patients with infected necrosis, the overall mortality rate had a tendency to decrease reaching 19% for the whole period. The best treatment results were achieved in year 2006 when overall mortality dropped to 12%. The relatively longer ICU stay in period 2 was related to a sizable increase of the necrotising forms compared to period 1. However, successful results of the conservative treatment and selective indications for surgical interventions resulted in shorter hospital stay with reduction from average of 25.0±23.2 days in period 1 to 23.6±21.0 days in period 2 and decreased the overall mortality rate reaching 18% in period 2 compared to 21% in period 1 (main outcomes are shown in Table VI).

Discussion

Management of acute pancreatitis continues to be an important medical issue due to the changing but not improving environment, ethiopathologic consequences of increasing incidence of gallstone disease, increased alcohol consumption, overweight with hypertrigliceridaemia and other factors predisposing to disease. At the same time, the number of recent international SAP management guidelines highlighted main aspects of the treatment strategy applicable in General Surgery and Intensive Care setting of the multiprofile hospital. The aim of our study was analysis of the recent eight-year experience in the treatment of SAP in our institution, which belongs to the above-mentioned category of hospitals.

The main improvement of the treatment results was achieved following internationally accepted recommendations; however, some aspects of the management were different and will be discussed.

The main aetiologic factor in our patients was alcohol and gallstone disease in a proportion close to what is reported in the European survey from Hungary ¹². A more focussed assessment and growing experience could account for the increasing number of patients with necrotising SAP admitted in the last period; however, further studies should be performed towards recognition of the main tendency.

The natural course of the disease was similar to that described in the international literature ¹³ although CRP – marker of the acute inflammatory response reached peak values on average three days after the start of the treatment and had typical decline approaching day 7. Dynamics of the organ dysfunction calculated according to the SOFA score were close to that of CRP. Relatively higher peak values of both CRP and SOFA score were observed in the second period when more necrotising forms were treated supporting our previous experience ¹⁴. Our data gives some optimism regarding the possibility of achieving positive treatment results even later than after 48-hour period from the onset of the disease when complex ICU treatment and monitoring are provided from admission day. Experience collected in the second period completely support evidence that ICU management is advisable for SAP patients as early as possible ¹⁵. Although the widely discussed antibiotic prophylaxis is not recommended in some of recent international publications ¹⁶,¹⁷,¹⁸ majority of patients with severe course of the disease received initial antibiotic prophylaxis with Ciprinol and Metronidasol or Imipenem/Cilastin. Further antibacterial therapy was routinely corrected according to the cultured microorganism. More selective indications for antibiotic prophylaxis were applied in the recent period, with main focus on the derangement of the tissue perfusion, multiple localisations of the exudate, early organ dysfunction and IAH.

Protocol treatment had an impact on the infection structure. The overall infection rate was in the internationally reported range ¹⁵,¹⁹. At the same time, infection associated with drains and catheters due to early surgical or interventional therapy dramatically decreased in the second period after implementation of the protocol and more selective surgical strategy. As a result, the second period demonstrated a considerably lower infection rate.

We observed predominance of the Gram-positive flora (Enterococci, S. epidermidis) isolated from the infected pancreatic and peripancreatic necrotic tissue followed by Enterobacteriacea (K. pneumoniae, E. coli) and nonfermenters (P. aeruginosa, A. baumanii, Stenotrophomona maltophilia). Our finding resembles recently reported observations from Germany ²⁰.

Nutritional support is one of the major constituents of the conservative therapy protocol. We completely agree with evidence-based recommendation that enteral feeding should be preferred whenever possible ²¹,²². Jejunal and gastric feeding could be successfully applied in cases when tube feeding is indicated ²³. Recently, we have introduced the mode of EOF and it is a clinical routine in our Surgical and ICU Departments ²⁴. To our knowledge, other authors do not report similar experience.

CVVH is another treatment modality included on routine basis in our SAP management protocol. Implementation of the CVVH was started since year 1999 and at the present moment our experience covers over 150 patients who underwent early CVVH due to severe clinical course of acute pancreatitis. Authors from China share their experience with application of mainly high flux CVVH in the treatment of SAP ²⁵,²⁶,²⁷,²⁸. Blood purification therapy as a grade C recommendation is included in acute pancreatitis management strategy guidelines from Japan ²⁹. Prevention and treatment of the IAH, pulmonary and renal dysfunction as a consequence of progressive SIRS and MODS were the main indications for application of early low flux CVVH in our series ³⁰.

Measurement of the IAP through the urinary bladder was started in parallel with implementation of the CVVH. Accuracy of the IAP measurements increased when all recommendations regarding the correct measurement technique were followed ⁹. This method can be safely used for monitoring of SAP patients and in combination with SOFA scoring; biochemical markers of SIRS and infection (Il-6, CRP, Procalcitonin test) are reliable severity predictors ³¹ and can reflect effectiveness of the conservative therapy.

Early recognition and prevention of systemic complications are essential determinants for favorable outcome. High incidence of pulmonary complications early during the clinical course of SAP is well recognised ³²,³³. Pulmonary complications positively correlated with SIRS and IAH in our series, confirming tight pathophysiologic relationship. Stricter adherence to SAP management principles improved diagnostic accuracy of pulmonary complications. Incidence of revealed pleural exudates and pneumonia increased in the second period. Prompt pulmonary support is of great importance in the maintenance of the pulmonary function. Despite the fact that more necrotising forms were associated with increased incidence of pneumonia and pleural exudates, it did not increase the rate of pulmonary dysfunction in the second period. Recognition of the IAH and balanced fluid replacement therapy with support of CVVH resulted in relatively low need for ventilatory support. Compared to the data from the international reviews ³⁴,³⁵ considerably fewer patients were managed on ventilatory support in our series. Positive correlation between pleural exudate and application of CVVH indirectly supported evidence that the procedure was done in patients with marked SIRS. On the other side, lack of correlation between CVVH and infection demonstrates the safety aspect of the procedure.

Effective conservative treatment consequently resulted in significantly decreased number of surgical interventions in the second period demonstrating even higher success rate compared to recently reported data ¹⁹. Formation of postnecrotic pseudocysts was a natural outcome in patients with necrotising pancreatitis. We observed a relatively higher rate of formation of pseudocysts than previously reported ³⁶. The new treatment strategy was associated with insignificant increase in average ICU stay in the second period, when more patients with necrotising forms were treated; however, it did not prolong but rather decreased the average hospital stay. Finally, progress with the more conservative approach resulted in a tendency towards decrease in the overall mortality rate. Main outcomes in our series are well comparable with internationally reported data ¹⁵,³⁷,³⁸.

Conclusion

Well-designed SAP treatment protocol is a rational way how to implement international experience ranging from high volume centres to a multi-profile hospital setting. Early focused assessment towards recognition of potentially severe clinical forms of acute pancreatitis at admission gives additional chance for favorable outcome even in cases when more than 48 hours have passed from the onset of the disease. Commencement of the initial resuscitation therapy, organ support and monitoring of the physiologic response including measurements of the IAP is recommended to provide in ICU setting. Early application of CVVH in order to prevent development or progress of organ dysfunction and IAH could be rational in patients with marked SIRS. Timely recognition and prevention of pulmonary complications is essential due its high incidence in SAP. Selective approach for antibacterial prophylaxis and application of the interventional therapy are important conditions for prevention of infection and control of sepsis. EOF is feasible and effective and can be provided in majority of SAP patients.

Conservative protocol-based approach is a rational treatment strategy for the management of SAP and can be successfully implemented in the setting of the university hospital. Further studies are needed for better understanding of the physiologic response, selective application of interventional therapies and control of sepsis.