PROGNOSTIC SCORE FOR SEVERITY OF LEPTOSPIROSIS

: The case-fatality rate of leptospirosis rises above 50% if intensive treatment is delayed. Our purpose was to create a simple score that could be used quickly to identify patients at the greatest risk. Material and Methods: One hundred consecutive patients with leptospirosis treated in the Clinic of Infectious Diseases at University Hospital “Dr Georgi Stranski” - Pleven (1976-2018) were distributed retrospectively in groups with a non-severe and severe course (66 and 34, respectively). The once interesting prognosis was an eventual risk for the severe course. Data were analyzed using statistical software (IBM SPSS Statistics 19.0). Results: The multivariate analysis identified seven variables independent associated with severe course: age ≥ 60 years OR (95% CI):15.45 (12.21–19.7), hospital admission >4 days after the clinical onset OR (95% CI):4.18 (2.23–7.86; oligo/anuria OR (95% CI):10.95 (3.2–19.8); hypotension OR (95% CI):13.936 (6.5–19.9); pulmonary findings OR (95% CI):25.45 (17.5–32.4); arrhythmia OR (95% CI):15.5 (10.2–19.6), and hemorrhages OR (95% CI):21.6 (18.3– 29.7). The mentioned variables (awarded one point each) were used to generate a seven-point prognostic score. The risk for severe course increased incrementally. A score of two, three, four or five had a positive predictive value of 30%, 38%, 74% and 89%, respectively (95% CI:28–39, 32–43, 60 – 90, 70–92, respectively). A score of six or seven had a positive predictive value of 100% (95% CI: 96–100 and 98–100, respectively). Conclusion: A simple seven-point clinical based prognostic score would help clinicians identify people at risk of severe leptospirosis and initiate prompt intensive treatment, including hemodialysis.


INTRODUCTION
Leptospirosis is a zoonosis with a worldwide distribution [1,2]. Although most cases of leptospirosis are mild and self-limiting, almost 60,000 people are fatal annually [1]. Severe leptospirosis manifesting as pulmonary haemorrhage, acute kidney injury (AKI) or multiorgan failure develops in 5-15% of cases. The case-fatality rate can rise to greater than 50% if the initiation of intensive treatment is delayed [3].
In everyday practice, identifying patients at risk of severe course can be difficult. According to different studies, the presence of a variety of clinical features, laboratory investigations and imaging and electrocardiography findings can be helpful [4,5,6,7,8,9]. These approaches may have less utility in low-resource countries, which bear a disproportionate burden of the disease [1, 10, 11].
Leptospirosis is not common in Bulgaria. The mean annual incidence in the Pleven region is 0.37 in 100 000 population within a 43-years period and correlates with the incidence for the country.
To improve the management of patients with leptospirosis and identify patient characteristics that predict a lethal outcome, we analyzed the presentation of cases with confirmed leptospirosis in the Pleven region and correlated their clinical findings and laboratory results with their subsequent clinical course. Our aim was to create a simple score that could be used to quickly identify the patients at greatest risk and begin immediately intensive treatment and support.

MATERIAL AND METHODS Study population and data collection
This retrospective study was performed in the Clinic of Infectious Diseases at University Hospital "Dr Georgi Stranski" -Pleven and recruited one hundred consecutive patients with leptospirosis from 1976 to 2018.
Patient characteristics at the time of presentation to medical attention were reviewed. Clinical findings, haematology, biochemistry, urinalysis, chest x-ray and electrocardiogram results were recorded. The data were obtained through the medical documentation of cases.
According to our previous study, severe leptospirosis had been defined at severe intoxication, intensive jaundice with severe hepatic dysfunction, skin hemorrhages and visceral bleeding, toxic myocarditis, severe ARF requiring dialysis, common respiratory and CNS involvement [12].
Based on the mentioned above definitions, the cases were distributed in two groups -with non-severe course (n1 = 66; 66% of all cases N = 100) and with severe course (n2 = 34; 34% of all cases N = 100). We avoided separate definitions for a mild and moderate course and considered the sum of them defining it as a non-severe course. The reason for this is the fact that during the first visit of a patient, it is difficult to distinguish mild and moderate severity. At the same time, the once interesting prognosis is an eventual risk for the severe course. https://doi.org/10.5272/jimab.2022282.4423 Data were entered into an electronic database (Microsoft Excel v. 2010) and analyzed using statistical software (IBM SPSS Statistics 19.0).
Moreover, these variables were analyzed by t-test, Ìann-Whitney's and χ 2 test (for parametric and non-para-metric distributions, respectively; p<0.05 was considered as significant); Pirson's test (ϕ-coefficient: weak correlation in ϕ<0.3, moderate -0.3<ϕ<0.7 and strong in ϕ>0.7); odds ratio (OR). For the multivariate analysis by linear and logistic regression, only variables with an area under the receiver operating characteristic (AUROC) curve of >0.7 in univariate analysis were selected.

RESULTS
During the studied period, one hundred patients with leptospirosis were treated in the Clinic of Infectious Diseases (87% of them were adults). The median age (interquartile range, IQR) was 35 years (22-57 years), and 90% were males. Urban residents were 61% (p>0.05). Summer seasonality was 78% (p<0.05). Risk exposures were registered in 88%, and half of these cases had an occupational hazard for leptospirosis. The disease had a severe course in 34%; a lethal outcome in 13%. Leptospirosis was the first clinical diagnosis of sixty patients. Comorbidity was registered at 34% and was more frequent in severe cases (21/34, 62%) compared with moderate cases (13/39, 33%) (p<0.0001). There was no comorbidity in mild cases.
Clinical characteristics were similar to those in the literature -jaundice and oligo/anuria (in 63%, respectively), conjunctival injection (87%), and hypotension (49%). The variation analysis revealed that respiratory symptoms, oligo/anuria, arrhythmia and hemorrhages were the commonest in severe cases (Table 1). Increased nitrogen parameters and serum bilirubin were with the highest levels at the severe course, and hypoproteinemia and hypoalbuminemia were with the lowest levels in the same group (p<0.0005) ( Table 2). Among thirty four severe cases, respiratory symptoms were observed in fourteen, hypotension in eight and hemodialysis in eight was performed. Thirteen patients were with lethal outcome after a mean of 4±2.6 days of hospital treatment (from 1 to 10 days). All of the deceased cases had respiratory symptoms (OR 25.0; p<0.0005).
The clinical onset of lethal cases was meanly five days before admission. Fever, myalgia, oligo/anuria were observed in all of them and epistaxis and hemorrhagic rash in two. Comorbidity was registered as follows: hypertonic disease and alcoholism in three, respectively, past myocardial infarct, stomach ulcer, past tuberculosis in two, respectively, and podagral and calculous cholecystitis in one, respectively. Six lethal cases were initially admitted to internal, and surgery wards with misdiagnosis (acute pancreatitis, obstructive jaundice, sepsis) and the adequate treatment was delayed. Acute renal failure in all, mental impairment and multiorgan bleeding in nine lethal cases were registered. The major factors leading to death were pulmonary and brain edema (OR 25.00; ϕ=0.659 and OR 17.29; ϕ=0.528, respectively) due to severe ARF. Moderate correlation of the death with ARF had found (OR 2.200; ϕ = 0.420), visceral bleed-ing (OR 1.43; ϕ=0.375), age more than 45 years (OR 11.00; ϕ = 0.43) and weak correlation with jaundice (OR 1.15; ϕ=0.031).
The multivariate analysis identified seven independent variables associated with the severe course. The first one was the age ≥ ≥ ≥ ≥ ≥60 years (OR (95% CI): 15.45 (12.21-19.7), p<0.0001), the second was the time of hospital admission more than four days after the clinical onset (OR (95% CI)  The mentioned seven variables -awarded one point each -were used to generate a seven-point prognostic score (Table 3).  Figure 3 is shown the distribution of non-severe and severe leptospirosis cases according to the score 0 to 7. It is obvious that all cases with score 0-1 were non-severe and all with score 6-7 -severe. All lethal cases had a score 6-7. The risk for severe course increased incrementally with the score. A score of one had a positive predictive value of 0%. A score of two, three, four or five had a positive predictive value of 30%, 38%, 74% and 89%, respec-tively (95% CI: 28-39, 32-43, 60 -90, 70-92, respectively). A score of six or seven had a positive predictive value of 100% (95% CI: 96-100 and 98-100, respectively) ( Table 4).

DISCUSSION
In patients with leptospirosis, a simple seven-point clinical score appears to reliably identify patients at risk of severe disease. The score could be used anywhere that leptospirosis is seen as a rapid and inexpensive assessment, which can be performed at the bedside of the patient. An absence of hypotension, oliguria, abnormal auscultatory findings, arrhythmia and hemorrhages is particularly helpful in identifying low-risk patients. The likelihood of severe disease rises incrementally as the score increases, facilitating recognition of the high-risk patient who needs prompt intensive treatment. Although many variables have been shown to pre-dict severe leptospirosis, a simple scoring system to quantify the risk of severe disease has proven elusive [5]. As in this series, older age has been associated with severe leptospirosis and worse outcomes in countries including India [6], Brazil [13] and Turkey [14]. Other predictors of severe leptospirosis have consisted of a combination of clinical features and laboratory findings. It is notable that pulmonary involvement is associated with worse outcomes in almost every published series, while renal involvement and hypotension also have been shown to have significant prognostic value. In the French West Indies, dyspnea, oliguria, white blood cell count >12,900/mm3, alveolar infiltrates on chest X-rays, and repolarization abnormali- The entirely clinical seven-point score has significant support for the clinical practice. It is simple to perform and emphasizes the renal impairment with consequent pulmonary and mental involvement that repeatedly have been shown to be associated with the worst clinical outcomes [9,21]. ARF in leptospirosis occurs due to direct leptospiral invasion resulting in tubulointerstitial nephritis [22,23]. Renal biopsy most frequently reveals a mononuclear cellular infiltration and interstitial edema, although immune-complex glomerulonephritis may also be present [21,22]. While leptospirosis has traditionally been thought to cause non-oliguric ARF [24], oliguria is an early clinical marker of ARF that is less likely to respond to rehydration [25,26]. Hypotension in leptospirosis is usually due to releasing of vasodilatory mediators and proinflammatory cytokines in response to the infection. The result is reduced renal blood flow, followed by renal injury [27]. Pulmonary involvement is one of the most serious manifestations of severe disease [24]. Leptospirosis impairs the maintenance of a fluid balance of alveolar epithelial cells resulting in pulmonary edema, which can trigger acute respiratory distress syndrome [28,29,30,31,32,33]. Pulmonary hemorrhage -the most dangerous respiratory manifestation -is thought to occur from a direct effect of leptospiral proteins or toxic cellular components upon the alveolocapillary membrane [34]. When areas of hemorrhage rise, the clinical signs are more likely to be apparent on auscultation [35]. As pulmonary hemorrhage progresses, vascular resistance in the lungs also increases, further contributing to systemic hypotension [27].
Severe disease was observed in one-third of our series, and the case-fatality rate was proportional. The wide variation in case fatality rates reported in the literature has been attributed to different definitions of severe leptospirosis. It is possible that the local case-fatality rate is higher than reported as the patients with rapidly fatal leptospirosis may have a negative MAT test during the early phase of the disease. As culture and PCR are not routinely used, clinical assessment has great importance. The excellent outcomes are in result of early disease recognition and access to prompt ICU support.
Our study was retrospective, and the seven-point score requires prospective validation. In conclusion, the simple seven-point clinical based prognostic score would help clinicians identify people at risk of severe leptospirosis. The score has the potential to improve the care of people with leptospirosis, particularly in resource-limited regions where the disease has a serious clinical burden.