Introduction:
Thousands of South African (SA) patients undergo surgical procedures each week and this population is not exempt from post-operative cardiovascular complications such as myocardial infarction, stroke, and cardiovascular mortality.(1-4) Peri-operative cardiovascular risk stratification in the non-cardiac surgery patient is performed using scoring systems based on the presence or absence of a series of risk factors, and yields the probability of post-operative cardiovascular complications.(2) These scoring systems include Lee’s Revised Cardiac Risk Index (RCRI).(5) Risk factors comprising the RCRI are ischaemic heart disease, heart failure, diabetes, stroke, renal impairment, and major surgery.(2) The information gained from the application of the RCRI can guide peri-operative decision-making and peri-operative risk reduction efforts.(2) The inclusion of specialised biochemical tests, such as pro-NT BNP, in pre-operative cardiovascular risk stratification has been shown to improve the prognostic accuracy of the overall risk stratification process when applied in well-resourced settings.(5,6) However, the use of specialised laboratory testing does have disadvantages in SA public health settings, where there might be limited access to specialised diagnostic resources.(7)
The use of a cheap, ubiquitous, and routinely performed laboratory test would be of great interest to those monitoring the peri-operative outcomes in resource-constrained settings such as South Africa. In SA settings, platelet counts are one of the laboratory tests which fulfil the criteria of being cheap, ubiquitous, and routinely performed. The American study of Glance et al.(8) reported a higher incidence of post-operative cardiovascular morbidity and mortality in non-cardiac surgery patients with platelet counts falling outside the upper and lower laboratory reference limits for this test (i.e. thrombocytosis and thrombocytopenia as defined using American laboratory reference ranges).
The findings of Glance et al. have not yet been confirmed in SA settings. Confirming what is known about the association between platelet counts and post-operative cardiovascular morbidity and mortality from overseas non-cardiac surgery settings in a SA setting will be the first step in establishing whether platelet counts should be considered as additional components of the peri-operative cardiovascular risk stratification inventory in SA settings. This was the purpose of our study.
Methods
Study Ethics:
This study was approved by the University of KwaZulu-Natal Biomedical Research Ethics Committee (BE015/18) which had also waived the requirement for individual consent due to the study being a sub-analysis of a retrospective registry.
Study design, patients and setting:
This was a propensity matched case-control study of adult vascular and general surgery patients who attended the quaternary-level Inkosi Albert Luthuli Central Hospital (IALCH) in Durban, South Africa between 01 January 2012 and 31 July 2016. The non-cardiac surgery specialties of vascular and general surgery were selected as these have been shown to be associated with high levels of post-operative morbidity and mortality in SA settings.(9)
Data, matching, and outcome definition:
The patients that constituted the study sample for our study were from a pre-existing surgical registry. Patients had data related to age, gender, individual RCRI risk factors (ischaemic heart disease, heart failure, stroke, pre-operative serum creatinine > 2.0 mg/dL, emergent surgery, and major surgery: supra-inguinal vascular/intra-peritoneal,/intra-thoracic surgery), pre-operative platelet count, and post-operative outcomes collected as part of the prior surgical registry. Platelet counts in this study were classified as normal, thrombocytopenia, or thrombocytosis based on the updated SA reference ranges provided by Lawrie and colleagues.(10) In this article the published platelet count reference ranges for men are 171–388 X109/L and for females 186–454 X109/L. We decided to use the SA reference ranges for platelet counts in this study as there are ethnic differences in platelet counts. Considering the obvious difference in the ethnic composition between SA and American populations, it would have been inappropriate to use the platelet references ranges from the study of Glance et al., in our study. In addition, the cumulative RCRI score was computed for each patient. The study outcome was post-operative major adverse cardiovascular events (MACE). This was defined as inpatient myocardial infarction, stroke or death following surgery. Post-operative myocardial infarction and stroke at IALCH were defined according to conventional definitions.(11,12) Post-operative survival was established from patient discharge summaries. The post-operative MACE outcome has been used in similar studies conducted in South African settings.(2,13) Cases were defined as patients who suffered post-operative MACE, and controls were defined as patients who did not suffer post-operative MACE. Propensity scores were computed for each patient in the original surgical registry based on the presence of individual RCRI risk factors. A sample size calculation indicated that a total of 150 patients at a case:control ratio of 1:1 would be adequate for this study. Patients were matched based on similar propensity scores until the required sample size was attained. Patients were not matched on age, gender, cumulative RCRI score, and platelet count category.
Data analysis:
We followed recommended guidelines for the statistical analysis of matched case-control data. Data was first analysed using univariate statistical methods. This included the McNemar test or McNemar-Bowker test for matched categorical data, and the paired t-test for matched continuous data. Results for the univariate statistical analysis are presented as frequencies and percentages, or means with standard deviations (SD), along with a corresponding p-value where applicable. A subsequent multivariate statistical analysis was performed wherein the unmatched characteristics (gender, age, RCRI score, and platelet count category) were then entered into a conditional logistic regression model. Results from the multivariate statistical analysis are presented as odds ratios (OR), with 95% confidence intervals (CI) and corresponding p-values where applicable. When interpreting the results of all statistical analyses, a p-value <0.050 was considered a statistically significant result. All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 24.0 (IBM Corp, USA).
Results
Description of the study sample:
Table 1 provides a description of the study sample. The mean age was 55.0 years (SD: 16.0 years). A total of 80/150 patients were male (53.3%). The most common RCRI risk factors were major surgery (80/150, 53.3%), diabetes (44/150, 29.3%) and ischaemic heart disease (37/150, 24.7%). Abnormal pre-operative platelet counts were observed in 25.3% (38/150) of all patients. Of the patients with abnormal pre-operative platelet counts, 12/38 (31.6%) had thrombocytopenia and the remainder (26/38, 68.4%) had thrombocytosis. The mean RCRI score for the study sample was 1.2 (SD: 0.7) points.
| Table 1: Description of the combined groups (N=150) |
| Characteristic |
Level |
n (% of N) or Mean (SD) |
Age |
N/A |
55.0 (16.0) |
Gender |
Male |
80 (53.3) |
|
Female |
70 (46.7) |
Diabetes |
Yes |
44 (29.3) |
|
No |
106 (70.7) |
History of ischaemic heart disease |
Yes |
37 (24.7) |
|
No |
113 (75.3) |
History of heart failure |
Yes |
1 (0.7) |
|
No |
149 (99.3) |
History of stroke |
Yes
No |
8 (5.3)
142 (94.7) |
Pre-operative renal impairment |
Yes |
5 (3.3) |
|
No |
145 (96.7) |
Major surgery |
Yes |
80 (53.3) |
|
No |
70 (46.7) |
Platelet count |
Thrombocytosis |
26 (17.3) |
|
Thrombocytopenia |
12 (8.0) |
|
Normal |
112 (74.7) |
Mean RCRI score (SD) |
N/A |
1.2 (0.7) |
SD: Standard deviation; N/A: Not applicable; RCRI: Revised Cardiac Risk Index |
Results of the univariate statistical analysis:
Table 2 shows the results for the univariate statistical analysis. As expected, none of the matched characteristics (i.e. the individual RCRI risk factors) were significantly different between the case and control groups. There was a statistically significant difference (p<0.001) in age between cases (Mean age: 59.4 years, SD: 14.8 years) and controls (Mean age: 50.6 years, SD: 16.0 years). Gender was not associated with post-operative MACE (p=0.766). Overall, a statistically significant univariate result was obtained across categories of platelet count (p=0.046). Furthermore, there was no statistical difference (p=0.567) between the mean RCRI scores of cases (Mean RCRI score: 1.2, SD: 0.7) and controls (Mean RCRI score: 1.2, SD: 0.7).
| Table 2: Results of the univariate statistical analysis |
| Characteristic |
Level |
Cases |
Controls |
p-value |
|
|
(n=75) |
(n=75) |
|
|
|
|
|
|
Mean age in years (SD) |
N/A |
59.4 (14.8) |
50.6 (16.0) |
<0.001 |
Gender |
|
|
|
0.766 |
|
Male |
42 (56.0) |
39 (52.0) |
|
|
Female |
33 (44.0) |
36 (48.0) |
|
Diabetes |
|
|
|
0.683 |
|
Yes |
22 (29.3) |
22 (29.3) |
|
|
No |
53 (70.7) |
53 (70.7) |
|
History of ischaemic heart disease |
|
|
|
0.999 |
|
Yes |
19 (25.3) |
20 (26.7) |
|
|
No |
56 (74.7) |
55 (73.3) |
|
History of heart failure |
|
|
|
0.999 |
|
Yes |
0 (0.0) |
1 (1.3) |
|
|
No |
75 (100.0) |
74 (98.7) |
|
History of stroke |
|
|
|
0.480 |
|
Yes |
4 (5.3) |
4 (5.3) |
|
|
No |
71 (94.7) |
71 (94.7) |
|
Pre-operative renal impairment |
|
|
|
0.999 |
|
Yes |
2 (2.7) |
3 (4.0) |
|
|
No |
73 (97.3) |
72 (96.0) |
|
Major surgery |
|
|
|
0.617 |
|
Yes |
41 (54.7) |
39 (52.0) |
|
|
No |
34 (45.3) |
36 (48.0) |
|
Platelet count |
|
|
|
0.046 |
|
Thrombocytosis |
10 (13.3) |
16 (21.3) |
|
|
Thrombocytopenia |
10 (13.3) |
2 (2.7) |
|
|
Normal |
55 (73.4) |
57 (76.0) |
|
Mean RCRI score (SD) |
N/A |
1.2 (0.7) |
1.2 (0.7) |
0.567 |
SD: Standard deviation; N/A: Not applicable; RCRI: Revised Cardiac Risk Index |
Results of the multivariate statistical analysis:
The results of the multivariate statistical analysis are shown in Table 3. After entering all the unmatched variables into a conditional logistic regression model, we found pre-operative thrombocytopenia to be independently associated with a six-fold higher odds of post-operative MACE when compared with normal pre-operative platelet counts (Odds Ratio: 6.28, 95% Confidence Interval: 1.19-33.12; p=0.031). The odds of post-operative MACE were similar between patients who had pre-operative thrombocytosis and patients who had normal pre-operative platelet counts (Odds Ratio: 0.78, 95% Confidence Interval: 0.31-1.93; p=0.585).
| Table 3: Results of the multivariate statistical analysis |
| Characteristic |
Level |
OR (95% CI) |
p-value |
|
|
|
|
Age (per year increase) |
N/A |
1.04 (1.02-1.07) |
0.001 |
Gender |
Male |
0.96 (0.49-1.90) |
0.914 |
|
Female |
Reference |
- |
Platelet count |
Thrombocytosis |
0.78 (0.31-1.93) |
0.585 |
|
Thrombocytopenia |
6.28 (1.19-33.12) |
0.031 |
|
Normal |
Reference |
- |
RCRI score (per point increase) |
N/A |
0.76 (0.46-1.24) |
0.275 |
OR: Odds ratio; CI: Confidence interval; N/A: Not applicable; RCRI: Revised Cardiac Risk Index |
Discussion
We found pre-operative thrombocytopenia to be an independent risk factor for post-operative MACE. Pre-operative thrombocytosis was not independently associated with post-operative MACE in our study. There are two proposed pathophysiological mechanisms for peri-operative cardiovascular complications. These are coronary vascular occlusion and prolonged imbalances of oxygen supply-demand in the myocardial tissue. Thrombocytosis facilitates the formation of coronary plaques, which can be disrupted during the peri-operative period and occlude the coronary vasculature. Thrombocytopenia predisposes patients to increased bleeding, hypovolaemia, and subsequent hypotension during the peri-operative period. The reduced coronary perfusion modulated by hypotension results in ischaemic myocardial injury.
From our observations, it seems likely that prolonged imbalances of oxygen supply-demand in the myocardial tissue through thrombocytopenia and increased bleeding is the predominant pathophysiological mechanism of post-operative MACE in SA non-cardiac surgery patients. As such, studies investigating the prognostic accuracy of supplementing the RCRI with pre-operative platelet counts for peri-operative cardiovascular risk stratification in SA settings are warranted. Our study also highlights the potential importance of addressing pre-operative thrombocytopenia in SA non-cardiac surgery settings. Evidence suggests that pre-operative platelet transfusions have no significant impact on peri-operative bleeding.(15) Estcourt et al.(16), recommend that other strategies such as artificial platelet substitutes, cryosupernatant/recombinant Factor VIIa/fibrinogen, desmopressin, thrombopoietin mimetics, recombinant Factor XIII or anti-fibrinolytics etc. might be more appropriate management options for pre-operative thrombocytopenia. The overall feasibility of these options in SA non-cardiac surgery settings is yet to be fully established, necessitating further research on this topic.
Study limitations:
Our study had limitations. Firstly, this study involved data from a single quaternary-level referral hospital. Therefore, our findings might not be generalizable. Secondly, we reported on inpatient outcomes, as post-discharge outcomes were not collected for patients in the pre-existing database. This is a common limitation of retrospective studies. We also know that HIV is associated with thrombocytopaenia in various stages of the disease process, however we did not include HIV status of the study population in data collection and subsequent analysis. Lastly, there was heterogeneity in the reference ranges for platelet counts used between our study and the study of Glance et al., which would have rendered any direct comparisons between the two studies inappropriate.
Conclusion:
We found thrombocytopenia to be independently associated with higher odds of post-operative MACE in SA non-cardiac surgery patients. No association between thrombocytosis and post-operative MACE was observed. These results suggest that prolonged imbalance of oxygen supply-demand in the myocardial tissue is the predominant pathophysiological mechanism of post-operative MACE in SA non-cardiac surgery patients. Pre-operative thrombocytopenia should be investigated as a potential component of future, peri-operative cardiovascular risk stratification systems in SA non-cardiac surgery settings. The feasibility of options for the management of pre-operative thrombocytopenia in SA non-cardiac surgery patients needs to be explored.
References:
- Botto F, Alonso-Coello P, Chan MT, et al. Myocardial injury after noncardiac surgery: a large, international, prospective cohort study establishing diagnostic criteria, characteristics, predictors, and 30-day outcomes. Anesthesiology. 2014;120(3):564-78.
- Moodley Y, Naidoo P, Biccard BM. The South African Vascular Surgical Cardiac Risk Index (SAVS-CRI): a prospective observational study. S Afr Med J. 2013;103(10):746-50.
- Biccard BM, Madiba TE, South African Surgical Outcomes Study I. The South African Surgical Outcomes Study: A 7-day prospective observational cohort study. S Afr Med J. 2015;105(6):465-75.
- Moodley Y, Biccard BM. The association between preoperative clinical risk factors and in-hospital strokes and death following carotid endarterectomy in South African patients. South Afr J Anaesth Analg. 2014;20(3):152-4.
- Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation. 1999;100(10):1043-9.
- Choi JH, Cho DK, Song YB, et al. Preoperative NT-proBNP and CRP predict perioperative major cardiovascular events in non-cardiac surgery. Heart. 2010;96(1):56-62.
- Petti CA, Polage CR, Quinn TC, et al. Laboratory medicine in Africa: a barrier to effective health care. Clin Infect Dis. 2006;42(3):377-82.
- Glance LG, Blumberg N, Eaton MP, et al. Preoperative thrombocytopenia and postoperative outcomes after noncardiac surgery. Anesthesiology. 2014;120(1):62-75.
- Moodley Y, Biccard BM. Predictors of in-hospital mortality following non-cardiac surgery: Findings from an analysis of a South African hospital administrative database. S Afr Med J. 2015;105(2):126-9.
- Lawrie D, Coetzee LM, Becker P, et al. Local reference ranges for full blood count and CD4 lymphocyte count testing. S Afr Med J. 2009;99(4):243-8.
- Tu JV, Wang H, Bowyer B, et al. Risk factors for death or stroke after carotid endarterectomy: observations from the Ontario Carotid Endarterectomy Registry. Stroke. 2003;34(11):2568-73.
- Mendis S, Thygesen K, Kuulasmaa K, et al. World Health Organization definition of myocardial infarction: 2008-09 revision. Int J Epidemiol. 2011;40(1):139-46.
- Biccard BM, Naidoo P, de Vasconcellos K. What is the best pre-operative risk stratification tool for major adverse cardiac events following elective vascular surgery? A prospective observational cohort study evaluating pre-operative myocardial ischaemia monitoring and biomarker analysis. Anaesthesia. 2012;67(4):389-95.
- Breslow NE, Day NE. Statistical methods in cancer research. Volume II--The design and analysis of cohort studies. IARC Sci Publ. 1987(82):1-406.
- Warner MA, Jia Q, Clifford L, et al. Preoperative platelet transfusions and perioperative red blood cell requirements in patients with thrombocytopenia undergoing noncardiac surgery. Transfusion (Paris). 2016;56(3):682-90.
- Estcourt LJ, Malouf R, Doree C, et al. Prophylactic platelet transfusions prior to surgery for people with a low platelet count. The Cochrane database of systematic reviews. 2017;2017(9):1-39.
|
