Anaesthesiology Intensive Therapy, 2011,XLIII,4; 172-177

Impact of the perioperative care model on mortality of patients treated in general surgery wards

*Mariusz Piechota


Department of Anaesthesiology and Intensive Therapy, Military University Hospital in Łódź

  • Table 1. The number of deaths recorded during hospital stay and mortality rates
  • Table 2. Transfers to intensive therapy units after prior hospitalisation in general surgery wards
  • Table 3. Relative mortality among patients with low death risk compared to patients of hospital III, based on the multiple regression model
  • Table 4. Patients with primary diagnoses of low death risk transferred to intensive therapy units after prior hospitalisation in general surgery wards
  • Table 5. Relative mortality among patients with moderate death risk compared to patients of hospital III, based on the multiple regression model
  • Table 6. Patients with primary diagnoses of moderate death risk transferred to intensive therapy units after prior hospitalization in general surgery wards
  • Table 7. Relative mortality among patients with high death risk compared to patients of hospital III, based on the multiple regression model
  • Table 8. Patients with primary diagnoses of high death risk transferred to intensive therapy units after prior hospitalisation in general surgery wards

Background. The risk of perioperative death in general surgery wards depends on many factors, including the underlying disease, type of surgical intervention and model of perioperative management. The aim of the study was to determine the reasons for major differences in mortality rates recorded in general surgery wards of the three university hospitals.

Methods. The retrospective study was carried out and involved the data of 32 231 surgical patients. In one of the hospitals, postoperative patients were treated in the recovery room supervised by anaesthetists; in the remaining two, perioperative care was delivered by surgical ward staff.

A multiple regression model with random effects was used to adjust for differences in three death risk groups of patients according to underlying diseases: low, moderate and high.

Results. In the hospital with postoperative care administered by anaesthetic staff the mortality rate was 0.45 whereas in the two remaining ones with postoperative patients supervised by surgical staff – 1.86 and 2.52. In each group, increased mortality was observed among patients receiving therapy in general surgery wards after transfer from another hospital ward.

Conclusion. The major factor determining the mortality rates in general surgery wards is the model of perioperative management.

According to statistical data, the surgery-related mortality is 1.4% [1]. The risk of perioperative death in general surgery is estimated at less than 1%. Mortality observed in surgical wards is a variable dependent on many risk factors, including, inter alia, the admission-related diseases, accompanying diseases, clinical condition, age, gender of patients, type of hospital admission, extent and type of surgical intervention, duration and type of surgical procedure, personnel skills and knowledge, medical errors, hospital organization and management, place of rendering, and sanitary-epidemiological conditions in the ward. Some of the risk factors mentioned, e.g. older age of patients [2] or urgent surgery [3, 4], have been well known for many years [5]. 

The aim of the present study was to determine the reasons for major differences in mortality rates recorded in general surgery wards and hospitals involved and to assess the effects of perioperative care models on mortality. The assumption made was that a decrease in mortality should be observed in patients treated in general surgery wards with higher perioperative care standards compared to those without such care. Moreover, it was assumed that patients admitted with the underlying diseases (primary diagnoses) statistically associated with higher mortality should benefit from higher standards of perioperative care compared to those with the underlying diseases of lower mortality.

METHODS

This retrospective study was approved by the Ethics Committee of the Medical University of Łódź and was based on the data of 32 231 patients treated in general surgery departments between January 1, 2003 and December 31, 2007 (according to dates of death or hospital discharge). In the period covered, there were three university hospitals in Łódź, all of which had at least one general surgery wards, i.e. the University Hospital no.1 (hospital I), no.2 (hospital II) and no. 5 (hospital III).

All general surgery wards and university hospitals included in the study were similar in a number of aspects. The hospitals were located at a short distance from one another. The surgical case-mix in general surgery wards was comparable. The majority of surgical procedures performed were abdominal and trauma surgeries. No cardiac, orthopaedic or neurosurgical procedures were carried out. In total, about 1200-2000 procedures a year were undertaken in each ward.

The studied data were obtained from the database of the Provincial Public Health Centre in Łódź in the form of individual electronic records of each patient. The analysed data regarded each hospitalisation in the surgical wards included as well as hospitalisations in other hospital wards following the first general surgery hospitalization.

At the first stage of study, the data obtained were processed to combine information relating to each patient covering the entire period of the patient’s hospital stay. The follow-up period commenced on the first day of hospitalization in the general surgery ward.

The assumption was that the basic factor determining mortality was the underlying condition (primary diagnosis) for admission to the surgery ward. Mortality was defined as the occurrence of death after the patient’s admission to the surgery ward, during hospitalization.

A multiple logistic regression model with random effects was used to adjust for differences in population distribution in the hospitals compared. The model factors were the patient age and sex, mode of referral and code of the underlying disease. All the variables, except for the code, were adopted in the model as fixed effects. Moreover, the significance of variable interactions was determined.

Separate models were adjusted to diseases (primary diagnoses) in three different risk groups: low, moderate and high. The first group – low death risk, included primary diagnoses with the level of mortality (with 95% CI) significantly lower compared to the average mortality level (for all hospitals). The next group – high death risk, included primary diagnoses with the level of mortality (with 95% CI) significantly higher compared to the average level of mortality (for all hospitals), i.e. 1.64%. The remaining primary diagnoses formed the group of moderate death risk.

The final model adopted for low death risk diseases did not incorporate the primary diagnosis as a confounding factor. The final model adopted for moderate death risk diseases did not include diseases (primary diagnoses) which caused hospitalization in fewer than 50 patients, which was related to a relatively small number of deaths observed for these causes of hospitalization and the fact that the causes occurred nearly exclusively in one of the hospitals compared.

Model of perioperative care in hospital III. In hospital III, the postoperative patients were treated in the recovery room supervised by the medical staff from the anaesthesiology and intensive therapy units. According to the model accepted, each surgical patient was provided with postoperative care based on the principles of ITU management until the attending anaesthesiologist decided that the patient can be transferred to an appropriate ward. The recovery room (a four-bed facility), located within the operating suite, was staffed with one anaesthesiologist and two anaesthesiology nurses providing continuous medical care to post-surgery patients. The ratio of anaesthesiologists and anaesthesiology nurses to the number of beds was 0.25 and 0.5, respectively. Surgeons acted as consultants throughout the recovery room stay and were responsible for typically surgical instructions.

The duration of stay in the recovery room depended on the condition of a patient, extent of surgery, coexisting diseases and possible complications. As a standard, patients stayed in the recovery room until the following postoperative morning (i.e. approximately 20-24 h).

The range of monitoring depended on the type and extent of surgery. Routine monitoring consisted of electrocardiography, indirect arterial pressure measurements, plethysmography, respiration rate, body temperature, fluid intake and diuresis. Extended monitoring was applied for major surgical procedures, such as low anterior resection or abdominoperineal resection of the rectum. Additional monitoring included central venous pressure in patients with a central catheter, direct arterial pressure measurements and capnography (in mechanically ventilated patients). In such cases, the recovery room stay was usually slightly longer.

Patients were managed in the recovery room until the attending anaesthetist decided otherwise. The model enabled early detection of complications and immediate institution of treatment. Another important advantage was the possibility to apply high-quality pain therapy. The regimen of pain management was based on a continuous fentanyl infusion (sometimes with non-steroidal anti-inflammatory drugs). The rate of infusion depended on the extent of surgical trauma, while the period of administration was determined by the duration of recovery room stay. The fentanyl infusion was stopped 0.5-1 h prior to transferring the patient to the surgery ward. The availability of infusion pumps enabled well-controlled and precise dosage of fluids and medications (e.g. catecholamines, nitroglycerine, β-blockers and hypotensive agents).

In life-threatening conditions (severe dysfunction or failure of one or multiple organs), the patient was referred directly from the operating suite or the recovery room to ITU for further treatment. Some high-risk patients were also hospitalized in ITU during the immediate preoperative period to provide optimal preparation for surgery.

Model of perioperative care in hospital I and II. Perioperative care in general surgery units of hospital I and hospital II was delivered by the surgical ward staff, i.e. surgeons and surgical nurses. The tasks of the team included proper preparation of the patient for surgery, early identification and treatment of life-threatening emergencies, early detection and treatment of acute postoperative complications, maintaining the homeostasis, provision of effective and appropriate postoperative analgesia. During the surgery, the patient was also under the care of the personnel of the anaesthesiology and intensive therapy unit.

Immediately after the surgical procedure, patients were transferred to the recovery room of the sending surgical ward. In some cases, they were sent to ITU, though not as frequently as in hospital III. The conditions, range of medical services and equipment provided in the recovery rooms of hospitals I and II meet the suitable requirements yet were much more modest compared to hospital III.

Statistical computations and analyses were performed using the logistic regression model with random effects. Statistical significance was assumed at p<0.05.

RESULTS

After the first stage of data processing the lowest number of death and the mortality rate were registered among the patients in general surgery ward of hospital III (Table1).

The largest group of patients transferred to ITU after prior hospitalization in general surgery unit was found in hospital III (Table 2).

Based on the multiple logistic regression model with random effects, no significant difference in mortality was noted in the low death risk group comparing hospital I, II, and III (Table 3). However, increased mortality rates were recorded among patients treated in general surgery wards after being transferred from another hospital department: OR 16.95 (95% CI: 3.75-76.59; p<0.001). The number of patients transferred to ITU after prior hospitalization in general surgery wards in the low death risk group  is listed in Table 4.

In the moderate death risk group, the analysis demonstrated significant differences in mortality between hospital III versus I and II (Table 5).  Increased mortality was also observed in  this group among patients receiving therapy in general surgery wards after being transferred from another hospital ward: OR 6.91 (95% CI: 3.91-12.23; p<0.001); patients aged 55 years or older: OR 1.74 (95% CI: 1.52-2.00; p<0.001) and among men: OR 1.57 (95% CI: 1.11-2.22; p=0.01). Data of patients transferred are listed in Table 6.

In the high death risk group, significant differences in mortality were observed between hospital I and II; no such differences were noted comparing hospital II and III (Table 7). However, increased mortality was recorded among patients treated in general surgery wards after transfer from another ward: OR 2.60 (95% CI: 1.92-3.52; p<0.001) and among patients aged 55 years or older: OR 1.51 (95% CI: 1.40-1.62; p<0.001). The number of patients transferred to intensive therapy units after the prior hospitalization in general surgery wards is listed in Table 8.

DISCUSSION

After adjusting for age, gender and mode of admission, no significant differences in mortality were noted among low death risk patients between hospital III versus I and II. The results showed that the factors such as knowledge and skills of medical staff, medical errors, organizational and management aspects, or sanitary and epidemiological conditions in the hospital wards were not significant causes of differences in mortality.

Furthermore, after adjusting for the non-homogeneous structure of primary diagnoses, age, gender and mode of admission, a significant difference in mortality was noted in the moderate death risk group, comparing the general surgery wards of the hospitals in question. The findings demonstrated that the mortality differences observed were not related to the structure of primary diagnoses, age, gender or mode of admission. Based on the results of low death risk group, it should be assumed that differences in the moderate death risk group were also not dependent on the knowledge and practical skills of the medical staff, medical errors, aspects organization and management, or sanitary and epidemiological conditions in a particular ward.

What factors, then, could be responsible for significant differences in mortality in the moderate death risk groups of the general surgery wards analysed? The significant difference in mortality was noted in patients who were not routinely transferred to ITU, unless specified otherwise. In this group of patients, the cause of significant differences in the mortality rate observed was related to the perioperative care model followed in the hospitals, which was supported by a proportionately higher number of patients from this group moved to ITU in hospital III in the perioperative period. Moreover, in hospital III, nearly all postoperative patients were treated in the recovery room supervised by the medical staff from anaesthesiology and intensive therapy units (in conditions and on principles equivalent to the ITU setting) or were transferred directly to ITU. This scheme of patient care is believed to produce the significant differences in mortality observed in the moderate death risk groups of the general surgery wards analysed.

After adjusting for the non-homogeneous structure of primary diagnoses, age, gender and mode of admission, a significant difference in mortality was noted in the high death risk group of patients between general surgery wards of hospital I and III. No such a difference was confirmed between general surgery wards in hospital II vs III. The findings in hospital I and III support the thesis that the perioperative care model determines differences in mortality. Moreover, the absence of a significant difference in mortality between general surgery wards in hospitals II and III does not contradict the thesis. In the high death risk group, the frequency of ITU hospitalizations was markedly higher in hospital II compared to other groups. Noteworthy, patients in the high death risk group were usually better monitored during their stay in the general surgery wards. Such patients normally stay in recovery rooms (run by general surgery unit staff) longer than patients from lower death risk groups do. This was likely to reduce substantially the difference in mortality in high death risk group of patients treated according to various perioperative care models and account for the lack of a significant difference in mortality between general surgery wards of hospitals II and III.

The study results show that the model of perioperative patient care was the key factor determining the mortality rate recorded in general surgery wards of three university hospitals in Łódź. Implementation of the perioperative care model followed at hospital III should significantly reduce the mortality of patients receiving treatment in general surgery wards. Moreover, a similar effect should also be observed in other surgery wards once they implement the hospital III perioperative care model. The model applied in hospital III is a far-reaching modification of the traditional perioperative care programme, developed to lower the mortality in general surgery wards. It is based on several simple, yet extremely important principles, including good cooperation between the general surgery wards and anaesthesiology and intensive care units, provision of postoperative care to all patients undergoing surgical procedures by anaesthesiology and intensive care staff at least during the first postoperative 20-24 h in conditions and on principles equivalent to the ITU setting. The fundamental difference between the hospital III perioperative care model and the traditional strategy was the possibility of appropriate monitoring of patients, adjusted to their clinical conditions, and the application of specialist knowledge and experience of anaesthesiology and intensive care personnel (e.g. in elderly patients with respiratory or cardiovascular impairment).

The effectiveness of the perioperative care model adopted in hospital III is validated by the following:

  • many postoperative complications are preventable by early identification of risk and appropriate treatment [6];
  • the highest number of postoperative deaths occur in elderly patients with coexisting heart or lung diseases undergoing major surgery [7];
  • the first 48 postoperative hours are critical for high death risk  patients [8];
  • elective admission of high death risk patients to ITU immediately after surgery should be seriously considered, as this procedure could substantially reduce postoperative mortality [8];
  • elevated mortality was observed among patients receiving non-optimal hospital treatment prior to ITU admission [9, 10];
  • the longer the hospital stay prior to ITU admission, the higher the mortality [11];
  • many surgical patients could benefit from ITU admission, but they are deprived of the opportunity [12, 13, 14];
  • appropriate monitoring and therapy of patients after surgical procedures could markedly lower mortality [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25];
  • some high death risk patients should receive elective ITU therapy during the postoperative period [8];
  • elderly patients should be admitted directly to ITU after surgery [36, 37];
  • with regard to the elderly, the National Confidential Enquiry into Perioperative Deaths  recommends better cooperation of surgeons, anaesthesiologists, medical practitioners and geriatricians [38].

However, it should be stressed, that the present study had some limitations: the data analysed do not describe the perioperative risk (e.g. ASA, POSSUM scores, etc). Nevertheless, it is believed that the findings presented should initiate further debates and lead to prospective, controlled studies comparing various perioperative care models.

CONCLUSIONS

1. The main factor determining the proportion of unsuccessful outcomes is the management of patients in the perioperative period.

2. The key aspect lowering mortality rates is continuous and meticulous specialist care, which offers the greatest benefits to patients in the group of primary diagnoses associated with a moderate death risk; high death risk patients are also expected to benefit significantly.

..............................................................................................................................................................

REFERENCES


1.    NHS Executive: Quality and Performance in the NHS. NHS Performance Indicators. London; 2000.

2.    Turrentine FE, Wang H, Simpson VB, Jones R: Surgical risk factors, morbidity, and mortality in elderly patients. J Am Coll Surg 2006; 203 : 865-677.

3.    Nilsson H, Stylianidis G, Haapamäki M, Nilsson E, Nordin P: Mortality after groin hernia surgery. Ann Surg 2007; 245: 656-660.

4.    Nilsson E, Kald A, Anderberg B, Bragmark M, Fordell R, Haapaniemi S, Heuman R, Lindhagen J, Stubberöd A, Wickbom J: Hernia surgery in a defined population: a prospective three year audit. Eur J Surg 1997; 163: 823-829.

5.    Suán C, Pérez-Torres C, Herrera A: Postoperative mortality in a general hospital. Rev Esp Anestesiol Reanim 1997; 44: 267-272.

6.    Gamil M, Fanning A: The first 24 hours after surgery. A study of complications after 2153 consecutive operations. Anaesthesia 1991; 46: 712-715.

7.    Buck N, Devlin HB, Lunn JN: The Report of a Confidential Enquiry into Perioperative Deaths. London: King’s Fund Publishing Office; 1987.

8.    Cavaliere F, Conti G, Costa R, Masieri S, Antonelli M, Proietti R: Intensive care after elective surgery: a survey on 30-day postoperative mortality and morbidity. Minerva Anestesiol 2008; 74: 459-468.

9.    McGloin H, Adam SK, Singer M: Unexpected deaths and referrals to intensive care of patients on general wards. Are some cases potentially avoidable? J R Coll Physicians Lond 1999; 33: 255-259.

10.    McQuillan P, Pilkington S, Allan A, Taylor B, Short A, Morgan G, Nielsen M, Barrett D, Smith G, Collins CH: Confidential inquiry into quality of care before admission to intensive care. BMJ 1998; 316: 1853-1858.

11.    Goldhill DR, McNarry AF, Hadjianastassiou VG, Tekkis PP: The longer patients are in hospital before intensive care admission the higher their mortality. Intensive Care Med 2004; 30: 1908-1913.

12.    Bennett-Guerrero E, Hyam JA, Shaefi S, Hyam JA, Shaefi S, Prytherch DR, Sutton GL, Weaver PC, Mythen MG, Grocott MP, Parides MK: Comparison of P-POSSUM risk-adjusted mortality rates after surgery between patients in the USA and the UK. Br J Surg 2003; 90: 1593-1598.

13.    Turner M, McFarlane HJ, Krukowski ZH: Prospective study of high dependency care requirements and provision. J R Coll Surg Edinb 1999; 44: 19-23.

14.    Green D, Paklet L: Latest developments in peri-operative monitoring of the high-risk major surgery patient. Int J Surg 2010; 8: 90-99.

15.    Pölönen P, Ruokonen E, Hippeläinen M, Pöyhönen M, Takala J: A prospective, randomized study of goal-oriented hemodynamic therapy in cardiac surgical patients. Anesth Analg 2000; 90: 1052-1059.

16.    Gan TJ, Soppitt A, Maroof M, el-Moalem H, Robertson KM, Moretti E, Dwane P, Glass PS: Goal-directed intraoperative fluid administration reduces length of hospital stay after major surgery. Anesthesiology 2002; 97: 820-826.

17.    Lopes MR, Oliveira MA, Pereira VO, Lemos IP, Auler JO Jr, Michard F: Goal-directed fluid management based on pulse pressure variation monitoring during high-risk surgery: a pilot randomized controlled trial. Crit Care 2007; 11: R100.

18.    Lobo SM, Salgado PF, Castillo VG, Borim AA, Polachini CA, Palchetti JC, Brienzi SL, de Oliveira GG: Effects of maximizing oxygen delivery on morbidity and mortality in high-risk surgical patients. Crit Care Med 2000; 28: 3396-3404.

19.    Poeze M, Takala J, Greve JW, Ramsay G: Pre-operative tonometry is predictive for mortality and morbidity in highrisk surgical patients. Intensive Care Med 2000; 26: 1272-1281.

20.    Montori VM, Devereaux PJ, Adhikari NK, Burns KE, Eggert CH, Briel M, Lacchetti C, Leung TW, Darling E, Bryant DM, Bucher HC, Schünemann HJ, Meade MO, Cook DJ, Erwin PJ, Sood A, Sood R, Lo B, Thompson CA, Zhou Q, Mills E, Guyatt GH: Randomized trials stopped early for benefit: a systematic review. JAMA 2005; 294: 2203-2209.

21.    Boyd O: Optimisation of oxygenation and tissue perfusion in surgical patients. Intensive Crit Care Nurs 2003; 19: 171-181.

22.    Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M: Early Goal-Directed Therapy Collaborative Group. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345: 1368-1377.

23.    Tote SP, Grounds RM: Performing perioperative optimization of the high-risk surgical patient. Br J Anaesth 2006; 97: 4-11.

24.    Goldhill DR: Preventing surgical deaths: critical care and intensive care outreach services in the postoperative period. Br J Anaesth 2005; 95: 88-94.

25.    Gal J, Bogar L, Acsady G, Kertai MD: Cardiac risk reduction in non-cardiac surgery: the role of anaesthesia and monitoring techniques. Eur J Anaesthesiol 2006; 23: 641-648.

..............................................................................................................................................................

Acknowledgements


The author wishes to acknowledge the help of Wojciech Sobala from The Nofer Institute of Occupational Medicine (NIOM) in Łódź who considerably contributed to this study.


..............................................................................................................................................................

adres/address:

*Mariusz Piechota

Department of Anaesthesiology and Intensive Therapy
Military University Hospital of Łódź
Haller Sq 1; 90-647 Łódź, Poland
tel: +48 42 639 30 70, fax: +48 42 639 30 97
e-mail: mariuszpiechota@poczta.onet.pl

otrzymano/received: 11.06.2011 r.
zaakceptowano/accepted: 20.08.2011 r.