ABSTRACT
Introduction: Acute kidney injury (AKI) requiring dialysis is an important health care burdenand is associated with very high in-hospital mortality. As no specific treatment is available toreverse AKI, the management remains supportive, including optimized fluid, electrolyte andacid-base balance, adjusting the dose of potentially nephrotoxic medications or avoidingsecondary haemodynamic and nephrotoxic kidney injury with timely initiation of dialysis.Timely initiation of dialysis in AKI is fundamental to achieve treatment goals and to providesolute clearance and removal of excess fluid while awaiting recovery of kidney function. Ifkidney function remains inadequate after a period of discontinuation from dialysis, it should bereinstituted by the foresight of the treating physician. The primary outcome of interest of thestudy was recovery of sufficient kidney function to discontinue haemodialysis therapy andcomplete recovery of renal function. Methods: This prospective observational study has beenconducted in the Department of Nephrology, Mymensingh Medical College Hospital,Bangladesh from September 2019 to February 2021. All adult patients in whom conventionalintermittent haemodialysis was initiated in the dialysis ward were included in the study. Allpatients were followed up till death or complete recovery or for a maximum period of six month.A structured data collection sheet was used to collect patients detail and recorded data wereanalyzed by IBM SPSS version 23. Results: A total of 134 patients of AKI requiring dialysiswere included in the study with the mean age of 42.3±15.7 years. Male (54.5%) were slightlymore than female with a male to female ratio of 1.2:1. Diabetes was present in 16 (11.9%)patients and hypertension was present in 47 (35.1%) patients. The causes of AKI weresepticaemia (35.1%), urinary tract infection (34.3%), surgery (18.7%), vomiting (16.4%),leptospirosis (11.2%), obstetric (10.4%), acute watery diarrhoea (9.7%), malignancy (8.2%), postrenal obstruction (8.2%), drugs (7.5%), rapidly progressive glomerulonephritis (6%), COVID-19(5.2%), rhabdomyolysis (4.5%), intestinal obstruction (3.7%), acute gastroenteritis (2.2%), waspbite (2.2%), organophosphorus compounds (OPC) poisoning (1.5%), starfruit toxicity (1.5%),haemolytic uremic syndrome (0.7%) and unknown (1.5%). Mean number of dialysis requirementwas 5.9±8.6 and length of hospital stay was 15.4±10.5 days. Out of 134 patients, 95 (70.9%)were discharged from hospital and 39 (29.1%) died in hospital. Total death of patients during thestudy period were 49 (36.6%) including home death of 10 (7.5%) patients. Complete recovery ofkidney function was achieved in 70 (52.2%) patients and partial recovery of kidney function whocan survive without dialysis were observed in 12 (9%) patients. 3 (2.2%) patients remain ondialysis and total survival during the study period was observed in 85 (63.4%) patients. Survivalrate was significantly higher in patients with ≤ 40 years (72.6%) and significantly lower inpatients with malignancy (18.2%) and post renal obstruction (27.3%). Conclusions: Outcomes ofacute kidney injury in patients requiring dialysis remains poor. Early detection, optimization offluid and electrolyte balance and timely initiation of haemodialysis are the keys to improvesurvival and overall mortality. No conflict of interest
ABSTRACT
Nowadays, Sentiment Analysis has become an active research area due to the availability of many opinionated data through increased activity in Blogging, Tagging, Podcasting, social networking sites, RSS feeds, and Social Bookmarking. In the present situation, the whole world is facing the crisis of the COVID-19 pandemic. Particularly, let's talk about nationwide lockdown in India to control the spread of COVID-19. The government relies on social media to observe people's aviews on their policies during the lockdown. In this paper, Twitter data has been used for Sentiment Analysis, which focus on people opinion during the COVID-19 nationwide Lockdown effect in India. Different keywords data was collected on various dates between March 25, 2020, to June 09, 2020. This research work is an application of the real-time TextBlob sentiment analyzer tool built based on the Natural Language Toolkit (NLTK). Relevant keyword tweets were extracted by tweeter API. Then a model was trained to classify the result on a specific opinion. This NLPbased sentiment analysis model is ideal for analyzing the emotions while tested with seven primary keywords: Lockdown1.0, Migrant Workers, Indian Economic, ICMR, Lockdown5.0, Medical Facilities, and Police. The result shows that Lockdown 1.0 got the most positive sentiments, followed by ICMR and Medical Facility. © 2021 IEEE.