regulations for MD General Medicine examination of the Tamil Nadu

(1)

To Study the long-term outcomes of patients with Acute Respiratory Distress Syndrome admitted to general medical wards and intensive care units in a tertiary care hospital in South India (L.A.R.S. Study)

A dissertation submitted in partial fulfilment of the rules and

regulations for MD General Medicine examination of the Tamil Nadu

Dr.M.G. R Medical University, Chennai, to be held in May 2020.

(2)

1

DECLARATION

This is to declare that this dissertation titled “To Study the long term outcomes of patients with Acute Respiratory Distress Syndrome, admitted to general medical wards and intensive care units in a tertiary care hospital in South India (L.A.R.S.

Study)” is my original work done in partial fulfilment of rules and regulations for MD General Medicine examination of the Tamil Nadu Dr. M.G.R. Medical University, Chennai to be held in May 2020.

CANDIDATE

Rohan Thomas Thomas

University registration number: 201711463 Post Graduate Registrar,

Department of General Medicine, Christian Medical College,

Vellore, Tamil Nadu.

(3)

2

CERTIFICATE

This is to certify that the dissertation entitled, “To Study the long term outcomes of patients with Acute Respiratory Distress Syndrome admitted to general medical wards and intensive care units in a tertiary care hospital in South India (L.A.R.S.

Study) ” is a bona fide work of Dr. Rohan T Thomas towards the partial fulfilment of rules and regulations for MD General Medicine degree examination of the Tamil Nadu Dr. M.G.R. Medical University, to be conducted in May 2019.

GUIDE

Dr. Thambu David Sudarsanam, MD, DipNB,

Professor and Head of Unit,

Department of General Medicine and Clinical Epidemiology, Christian Medical College,

Vellore.

(4)

3

CERTIFICATE

This is to certify that the dissertation entitled, “To Study the long term outcomes of patients with Acute Respiratory Distress Syndrome admitted to general medical wards and intensive care units in a tertiary care hospital in South India (L.A.R.S.

Study)” is a bona fide work of Dr. Rohan T Thomas towards the partial fulfilment of rules and regulations for MD General Medicine degree examination of the Tamil Nadu Dr. M.G.R. Medical University, to be conducted in May 2019.

PRINCIPAL HEAD OF THE DEPARTMENT

Dr. Anna Pulimood, Professor,

Department of Pathology, Christian Medical College, Vellore.

Dr. Thambu David Sudarsanam, Professor and Head,

Department of General Medicine and Clinical Epidemiology,

Christian Medical College, Vellore.

(5)

4

URKUND ANTIPLAGIRISM CERTIFICATE

This is to certify that this dissertation work titled “To Study the long term outcomes of patients with Acute Respiratory Distress Syndrome admitted to general medical wards and intensive care units in a tertiary care hospital in South India (L.A.R.S.

Study)” of the candidate Dr. Rohan Thomas Thomas with registration number 201711463 in the branch of General Medicine. I personally verified the urkund.com website for the purpose of plagiarism check. I found that the uploaded thesis file contains from introduction to conclusion pages and result shows 0% percentage of plagiarism in the dissertation.

GUIDE Dr. Thambu David Sudarsanam MD, DipNB Professor and Head of Unit Department of General Medicine and Clinical Epidemiology

(6)

5

ACKNOWLEDGEMENT

There have been many people who have been of great value in guiding, assisting and refining this thesis. I would like to acknowledge and be grateful for their part in the making of this thesis.

To my guide Dr. Thambu David Sudarsanam, Professor and Head of Department of General Medicine and Clinical Epidemiology, for his and the knowledge and principal of ethical research he has imparted. I thank him for his patience and kindness with which he dealt with through this dissertation.

To my co-investigator Dr. Vijay Prakash Turaka, for his guidance and input in the

making and refining my dissertation. I would also like to thank my other co-investigators Dr. D J Christopher, Dr. Balamugesh, Dr. Thomas Issiah Sudarsan, Dr. Alice Joan

Mathuram, Dr. Raamya I, Dr.Mohammed Sadiq and Dr. Tarun K George for all their valuable input my dissertation I would like to thank Mrs. Mahasampath Gowri and the Department of Biostatistics for their assistance in statistical analysis.

I am grateful to my colleagues in the department of medicine for their assistance in recruitment.

To Institute which was provided the platform for my dissertation to be possible.

I thank the patients and their relatives for not only consenting to participate in this study.

I would like to thank my family and friends for their constant support and encouragement.

(7)

6

Index of Tables

Table 1: Berlin Definition of A.R.D.S. ... 15

Table 2: Baseline characteristics ... 58

Table 3: Berlin criteria ... 66

Table 4: Severity of A.R.D.S. ... 68

Table 5: 6-month Mortality of A.R.D.S. survivors in our cohort ... 69

Table 6: Return to work of A.R.D.S. survivors* ... 70

Table 7A: SF 36 Quality of Life Questionnaire ... 71

(9)

8

Table 8 Comparison of patient co-morbidity profiles among Canadian cohort, Karnataka cohort and our cohort ... 81

Index of Figures

Figure 1Aetiology of A.R.D.S. in a south Indian Hospital in Karnataka ... 21 Figure 2: Mortality, 6-Minute Walk Distance, and Quality of Life scores for 5 Years after Discharge in the cohort by Herridge et al. ... 40 Figure 3: Median scores in the SF 36 quality of life questionnaires in the cohort by

Herridge et al. ... 41 Figure 4 Strobe Diagram ... 57

(10)

9

Acute Respiratory Distress Syndrome (A.R.D.S.), a syndrome described more than 50 years ago by Ashbough was a disease entity that has gain significant prominence over the past few years. In the past, there were very few survivors with a paucity of strategies to combat the severe respiratory failure. However, over the past few decades, and thanks to the work of many researchers into this illness, there has been a drastic improvement in mortality of patients with A.R.D.S. This is in part due to the use of newer ventilatory strategies, such as a low tidal volume ventilation strategy and other improvements in intensive care. Though there is still scope for research into further improving the short- term outcomes, due to the increased population of survivors, the long-term effects of the illness are slowly coming to light. A large cohort study done in a western population noted significant morbidity in survivors even up to 5 years after the illness. However, we have almost no data on the same in India and risk factors for these long term sequalae are yet to be clearly defined. As the first step to tackling any problem is to identify it, we decided to follow up our patients who had survived this illness. This study was done to assess the clinical profile, long term mortality, abnormalities in pulmonary function on spirometry, exertional capacity and overall economic burden of patients with acute respiratory distress syndrome admitted in the intensive care units and general medicine wards.

1 Introduction

(11)

10

To study the long term physical and quality of life outcomes of patients with Acute Respiratory Distress Syndrome who are admitted to the medical intensive care unit and general medical wards and in a tertiary care hospital in South India.

2 Aim

(12)

11

3.1 Primary Objective

To determine 6 months outcomes of patients with A.R.D.S.

3.2 Secondary Objective

a. To determine the mortality of patients with A.R.D.S., 6 months after discharge b. To calculate the percentage of patients with A.R.D.S. who return to work in the study period.

c. To determine the direct medical costs (hospital bill including medicines) as well as the indirect costs (wages lost during hospitalization as well as time to return to work)

d. To assess the quality of life reported by the patients after an episode of A.R.D.S 6 months after discharge.

e. To identify risk factors that may predict mortality and morbidity.

f. To assess the average distance walked on a the 6-minute walk test and outcomes of spirometry in survivors of A.R.D.S., 6 months after illness.

3 Objectives

(13)

12

4.1 Historical origins of acute respiratory distress syndrome

The entity we now call Acute Respiratory Distress Syndrome has existed far before the term was coined in 1967.One of the earliest descriptions was given by Dr. René Laennec, who after

inventing the stethoscope, described a condition of pulmonary edema without heart failure in 1816. A condition which at the time, was almost universally fatal.(1) Multiple names have been used for the term since then including shock lung, respiratory lung, double pneumonia and so on.

Acute Respiratory Distress Syndrome as a term was first coined 50 years ago by Dr. Ashbaugh and his colleagues. He had initially described a case series of 12 patients who had tachypnea, with refractory hypoxemia and bilateral opacities seen on a chest radiograph. It was again associated with a high mortality and there were no clear outlines for diagnosis or treatment. He described a variety of causes, from severe infections to burns which gave rise to this disease.

Now in 2017 a considerable amount of progress has been made regarding A.R.D.S.(2), with new definitions, management strategies and information regarding short term outcomes. Though the complex pathophysiology is not yet fully understood, the benefits of this knowledge are clearly seen in the change in mortality due to A.R.D.S, as seen in a large epidemiological study by Cochi et all in the U.S.A. The mortality rate was noted to decrease from 5.01 to 2.82 per 100,000 person years between the time frames of 1999 to 2013.(3)

4 Review of literature

(14)

13

As the mortality decreases, the number of survivors continues to increase, and a new problem arises. Though, there is a paucity of data on long term outcomes, internationally, a recent 5 year cohort study was completed , which showed significant physical disability and persisting impairment in the quality of life of patients recovering from A.R.D.S.(4). It was also noted that at one year, nearly 40 % of people who had been diagnosed with A.R.D.S. prior, had not

returned to work. Patients had diminished exercise and vital capacities persisting for years and not fully reaching their pre-A.R.D.S. expected baselines.

However, all this data only comes from studies on western populations. With regards to the long-term outcomes of A.R.D.S. in the Indian continent, there is currently very little data available.

4.2 Definition of A.R.D.S.:

There have been four major definitions of A.R.D.S. so far. The original description by Ashbaugh is present in all these definitions.

(15)

14

The most widely accepted definition till 2012 was the American-European Consensus

Conference (AECC) definition, which defined A.R.D.S. as sudden onset respiratory failure, with bilateral infiltrates (based on a chest radiograph), and hypoxemia (defined as a

PaO2/FiO2 ratio ≤200 mmHg), with no evidence of cardiogenic pulmonary edema, ( i.e. no left atrial hypertension or a pulmonary capillary pressure <18 mmHg (if checked) ). The

PaO2/FiO2 ratio between 200-300 was at the time known as an Acute Lung Injury.

However, there have been many drawbacks with the A.E.C.C. definition. When compared to the histopathological findings of diffuse alveolar damage, this definition was found to have a low specificity of only (51%) on autopsies performed on patients diagnosed with the above.(5). It was also noted that there was significant inter observer variability with regards to chest x ray findings, and the PaO2/FiO2 ratio varied with the Positive End Expiratory Pressure used in ventilation which was a variable not addressed by the definition. Also, the diagnosis of A.R.D.S.

and the presence of an elevated pulmonary wedge pressure are not mutually exclusive as was implied by the definition and even patients with heart failure may develop a concomitant acute respiratory distress syndrome. The above problems with the definition, led to an international expert panel meeting in 2011, and a new definition , the Berlin definition was made (6). This definition has superseded the previous American European Consensus Conference Definition.

(16)

15

The Berlin definition has been validated in terms of its ability to predict mortality(7) (based on a meta-analysis of 4188 patients) with a reasonable co-relation with histopathological

findings(8). The Berlin definition of A.R.D.S. is as follows:

Table 1: Berlin Definition of A.R.D.S.

Criteria Rationale

Onset within 7 days Observational studies showed ARDS develops within 72 hours- To 1 week

Bilateral opacities Opacities should not be fully explained by another single cause Category of severity (P/f ratio) Based on the above meta-data

Mild 201-30, mortality 27% (95% CI, 24-30)

Moderate 101-200, mortality 32% (9% CI 29-34) Severe <100, mortality 45% (95% CI 42-48)

(on a minimum PEEP of 5) To standardize P/F ratios. Also, oxygen delivery is reliable mainly with NIV and invasive ventilation

(17)

16

4.3 Global burden of disease:

A.R.D.S. is a very common problem in healthcare. The LUNGSAFE study (Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure) a large

international multicentric study done in 459 ICU, provides us a look at the extent of the problem.

10.4% of all ICU admissions( 29144 ICU patients screened) were noted to have A.R.D.S. as per the Berlin criteria (9). And as many as 23% of all patients admitted requiring mechanical

ventilation were noted to meet the BERLIN criteria for A.R.D.S. (10). Internationally reported incidences in the western population have ranged from 17-64 per 100,000 person-years and 17- 34 in European countries.

A.R.D.S. is seen in both the young and the old. However, in western literature the incidence appears to be more in the elderly, with 16 per 100,000 person-years in those between 15-19 years of age and 306 per 100,000 person-years among patients 75- 84 years.

In India, a prospective observational cohort study was done of all mechanically ventilated

patients in one year in a surgical ICU in Christian Medical College Vellore, and the incidence of A.R.D.S. was found to be 11.6 %.(11).

(18)

17

Magazine et al also did a retrospective analysis of all patients with Acute Respiratory Distress Syndrome in a tertiary hospital in Karnataka He noted that the mean age was 42 years and of the 150 patients with A.R.D.S ,53 % of them were below 40 years of age with 60% being males..

The mortality rate in the study was around 41.8%, like international studies. The mean duration of admission was 9.7+/- 5.5 days..(9)

Extrapolating the above data, with 10-20% of the patients in ICU having A.R.D.S. and a mean duration of ICU stay of 10-15 days, a typical 10 bed ICU hospital in Asia could see around 35 cases in a year

4.4 Pathogenesis:

The pathophysiology of A.R.D.S. is complex and still not fully understood. In the normal lung the alveoli are relatively dry, with various mechanisms to prevent fluid leak, which would impair gaseous exchange. The retained intravascular protein provides oncotic pressure, the interstitial lymphatics drain the excess fluid and tight junctions between alveolar epithelial cells prevent leakage contributing to keeping the alveoli dry.

The first trigger is an injuring stimulus which triggers a pro inflammatory response involving tumor necrosis factor, IL – 1, IL-6 and IL – 8. This causes neutrophil migration to the lungs and

(19)

18

release of proteases and reactive oxygen species from the activated neutrophils. This starts the so called inflammatory or exudative stage. The main characteristic is an increased capillary

permeability due to damage to the capillary and alveolar endothelium.(12). This causes protein rich fluid to accumulate in the capillary. The inflammatory cytokine response also later causes diffuse alveolar damage.

There are classically 3 described phases of ARDS, an exudative/inflammatory, a proliferative phase and fibrotic phase. Initially, during the first days, as described above, there is

accumulation of edema fluid into the lungs with macrophages, neutrophils and red blood cells seen in the alveoli. There is also epithelial and endothelial injury with hyaline membranes seen.

To repair the injury, there is proliferation of type 2 alveolar cells and fibroblasts, with new matrix deposition, and some fibrosis. This is called the proliferative phase. There is gradual resolution of inflammation and resorption of some of the edema fluid.

However, some patients develop significant fibrosis later, known as the fibrotic phase, and can lead to reduced pulmonary compliance. The development of the fibrotic phase is suggested to be associated with an increased mortality.(13) This represents the subset of patients who do not show dramatic improvement even after a week of ventilation.

The fibroproliferative response was initially thought to be a very late occurrence in the

pathophysiology of A.R.D.S. However newer hypothesis suggests that it begins even as early as

(20)

19

24 hours after the development of A.R.D.S. In a study done by Richard et al, patients with acute respiratory distress syndrome underwent a bronchoscopy and bronchial lavage at day 1 and day 7 of diagnosis. The fluid from the lavage was tested in vitro on human lung fibroblast cells and demonstrated the ability to stimulate their growth. This stimulation co-related to its

concentration of N- terminal peptide for type 3 procollagen, a marker which also appeared to predict mortality in the patient subsets. This suggests the pro-fibroblastic stimulus occurs even early in the disease and could be a predictive marker of mortality.(14)

Especially when dealing with long term morbidity in survivors, the fibroproliferative changes are one of the implicated pathologies for the persistent impairment of physical function and poorer quality of life. Some indirect evidence comes from a few small sized studies which compared quality of life and pulmonary function tests, noting a modest correlation between a lower FEV1 and FVC and poorer physical scores. (15). A retrospective analysis of prior studies showed a correlation with reticular nodular patterns on CT and a restrictive pattern on spirometry and DLCO. This overall suggests that fibrosis played a role in the persistent morbidity of ARDS.(16) The extent of fibrosis and the clinical significance of the same is still under exploration.

(21)

20

This could pave the way for many novel treatment strategies to mitigate long term mortality such as tyrosine kinase inhibitors to prevent fibroblastic proliferation.(17)

4.5 Etiology and risk factors:

There are a variety of etiologies for the development of A.R.D.S. The commonly known ones being pneumonia, sepsis, gastric content aspiration, trauma, pancreatitis, inhalation injury, burns, non-cardiogenic shock, drug overdose, acute lung injury following massive transfusions

(TRALI), drowning, etc.(6)

Sepsis is the most common cause of A.R.D.S.(18) followed by community acquired pneumonias, both bacterial and viral. A study on etiologies of pneumonia in A.R.D.S., showed the incidence of viral infections in community acquired pneumonia causing A.R.D.S. to be nearly 29%(19) . Other causes include aspiration pneumonia(where A.R.D.S. occurred in nearly a third of all patients), trauma, pancreatitis, burns, transfusions and rarely drugs.(20) The etiologies as

described by Magazine et all in his epidemiological study in a tertiary hospital in south India are as follows. (20)

(22)

21

Figure 1Aetiology of A.R.D.S. in a south Indian Hospital in Karnataka

(Figure taken from Magazine et all , Epidemiological profile of acute respiratory distress syndrome patients: A tertiary care experience, Lung India ,January 1^st 2017 (20))

Some risk factors associated with the development of A.R.D.S. were Cigarette smoking(21) and a high BMI.(22). This was based on a retrospective cohort in California of 121000 people with 55 cases of A.R.D.S. An association between cigarette smoking and risk of developing A.R.D.S.

was suggested with an odds ratio of 2.85 for those who smoked < 20 pack years vs nonsmokers and 4.59 for smokers of >20 pack years vs nonsmokers.

(23)

22

There was no co-relation noted between alcohol consumption and A.R.D.S. Similarly, for B.M.I, in a cohort of 1795 critically ill patients at risk for A.R.D.S., a strong odds ratio for its

development with increasing B.M.I. was noted. Possible explanations were the pro-inflammatory effects and free radical damage to lungs caused by smoking and increased inflammatory

adipokines noted in obesity.

Genetic markers have also been proposed to have an implication in A.R.D.S. This is based on the observation that , in a lot of patients, despite risk factors , A.R.D.S. does not seem to

develop(2).The reason why some patients develop A.R.D.S. and some do not is still under study, with many genes such as those coding for ACE, IL-10, VEGF, etc. showing some

association(23),(24).

4.6 Treatment:

As mentioned above, A.R.D.S. was nearly universally fatal during the time of Dr. Rene Laennec.

The first step towards improving mortality in patients with A.R.D.S. was the establishment of an intensive care unit by Dr. Bjørn Aage Ibsen. Along with this, first mode of mechanical

ventilation, the famous iron lung invented by Drinker and Shaw in 1929 was another step towards improving outcomes in A.R.D.S. However, unlike patients with polio who had very compliant lungs, those with A.R.D.S. were noted to have stiff lungs, which did not ventilate well

(24)

23

with the negative pressure concept used by the iron lung. Thus in 1959, Dr. Ibson created the first mode of positive ventilation, the cuffed endotracheal tube. This was the first step in changing what was a universally fatal disease to one with a fair chance of survival. Since then there has been a lot of progress into improving management and thereby reducing mortality in A.R.D.S.

Many new ventilatory strategies were developed to optimally address the severe hypoxemia associated with A.R.D.S. and minimize the complications of mechanical ventilation.

The large metacentric RCT, done in 2000, the famous A.R.D.S. Network study, looked at 861 patients in multiple centers. They found a statistically significant decrease in mortality with a low tidal volume strategy vs the conventional tidal volume strategy. (31% mortality vs 39.8%) (25)(26).

Another landmark trial for the low tidal volume strategy, the AIRES trial, done in eight centers in Spain, randomized a 103consecutive patients which met the A.E.C.C. criteria for A.R.D.S. at presentation and 24 hours after conventional ventilation, to a high PEEP low tidal volume strategy vs conventional ventilation, They found a statistically significant decrease in ICU

mortality (53.3%vs. 32%, p = .040) and ventilator free days at day 287(6.02 ± 7.95 days vs 10.90

± 9.45 days, in p = .008)) with the high PEEP low tidal volume strategy. This finding was significant enough to satisfy the early termination criteria for clear benefit with a >20%

(25)

24

decreased in mortality of one arm to the other (after a minimum of 45 patients were present in each arm to reduce alpha error to 2.5%).

Other measures of ventilation like prone ventilation in A.R.D.S. with a pulmonary function ratio of less than 150 has also shown a mortality benefit.(27). Another Landmark trial, the PROSEVA trial , a prospective randomized control trial with 466 patients with A.R.D.S. with PaO2/FiO2 ratios of less than 150 showed that in such patients, prone ventilation for at least 16 hours a day till improvement of P/F in supine position or 28 days had a profound impact in improving short term mortality, with a hazard’s ratio of 0.39 (95% CI: 0.25 to 0.63). Both the 28 and 90 day mortality were significantly less in the prone ventilated group.(28)

Use of muscle paralytics had initially been shown to reduced mortality and improve time-off ventilator.(29). This changed however, after a large RCT done by the The National Heart, Lung, and Blood Institute PETAL Clinical Trials Network. 1008 patients with A.R.D.S. with a

PaO2/FiO2 ratio of <150 on invasive ventilation with a Positive End Expiratory Pressure > 8 cm H20 were randomised to 48 hours of neuromuscular blockade vs routine sedation. It was noted that there was no significant difference in 90-day mortality and the trial was terminated early due to futility in the interim analysis.

(26)

25

As we have also come to learn, the most important treatment to hasten resolution of A.R.D.S. is to address the underlying etiology. Improvements in supportive care have significantly decreased the mortality of the associated hypoxemia and ventilator complications.(2)

Many years ago, Ashbough and colleagues had also suggested a probable beneficial role of glucocorticoids int eh management of A.R.D.S. However, this is still a topic of uncertainty, with data suggesting harm if given late in the disease. This was noted in a randomized control by Steinberg in 2006, involving 180 patients with A.R.D.S randomly assigned to a tapering dose of methylprednisolone vs control. He found a significant increase in mortality in patients given steroids if the same was given after 14 days. When given early in ARDS though, he noted more 28-day ventilator free days and a shorter hospital stay in the methyl prednisolone group.

However this findings have been inconsistent across other studies and clear evidence for or against the use of steroids is still needed.(30).

It had been suggested that a lower oxygen saturation target of 88-92 % vs a liberal target of >96

% was beneficial in terms of mortality and organ dysfunction.(31) This was supported by a study by Rakshit et all in 4 ICU centers, where a 108 patients requiring mechanical ventilation were randomly assigned to liberal vs conservative oxygen targets. The 90-day mortality was lower in the conservative group (when analyzed for the subgroup with a PaO2/FiO2 ratio <300) with a hazard’s ratio of 0.49 (95% CI:0.20–1.17). Another study by Pierce et all, assessing a

(27)

26

cohort of 2,994 A.R.D.S. patient who received excessive or above goal oxygen (defined as a PaO2 above 80 on an arterial blood gas) with those below the goal noted an increased mortality and longer duration of ventilation of those who received excessive oxygen. The proposed

mechanism was hyperoxia mediated Reactive oxygen Species damage and vasoconstriction of distal vessels. However larger RCT will be needed to confirm the association.

How these various treatment strategies affect the long-term morbidity and mortality of ARDS is unknown. As of now there are no clear guidelines for strategies to prevent long term morbidity in the context of A.RD.S.

4.7 Outcomes:

There has been a lot of literature on short term outcomes for A.R.D.S. The metanalysis used to formulate the Berlin definition had shown mortality rates as high as 45% in severe ARDS.

The initial few days of A.R.D.S. is characterized by severe hypoxemia needing supplementary oxygen and often mechanical ventilation with a high positive end expiratory pressure. This is followed by improvement in oxygenation allowing weaning of mechanical supports over the

4.7.1 Short term outcomes:

(28)

27

next few days. Some patients, however, may have a prolonged duration for resolution of ARDS due to progression to the fibroproliferative stage. This also may account for some of the long- term morbidity of this disease.

Literature currently suggests mortality rates of 34.9% in mild A.R.D.S., 40.3% in moderate A.R.D.S. and 46.1% in severe A.R.D.S. in a multicentric international epidemiology study in 2017.(9)

Looking at Indian data, in 2 tertiary centers in Karnataka and Mumbai, the mortality of patients with A.R.D.S. was around 44% and 58% respectively. (20,32)

There are a considerable number of factors that affect short term mortality in A.R.D.S.

Factors can be divided into

a) Patient related – Age, Obesity, Underlying co-morbidities

b) Disease related – Severity of A.R.D.S. APACHE/SAPS score. Etiology of the A.R.D.S.

c) Treatment related – Fluid Balance, Use of glucocorticoids, Transfusions and late intubations.

4.7.2 Factors affecting short term outcomes:

(29)

28

It was noted that mortality increased with increasing age, seen in a multicentric cohort of 1113 patients in Kings country with a mortality of 24% for age groups of 15 -19 years to 60% for those 85 years or older.(33).

With regards to obesity, studies are conflicting, with most suggesting that in obese patients with only respiratory failure, mortality is the same as the non-obese. However, if any other organ involvement is also present, the mortality appears to be higher for an obese patient than a matched non obese patient. This data had come from a western analysis of obese patients

undergoing invasive mechanical ventilation(34). Multiple other observational studies have also shown no increase in mortality with some even suggesting an inverse relationship, though this has not been established conclusively.(35–37)

With regards to disease severity , as mentioned in multiple studies , most notably the ones used to frame the Berlin definition , the more severe the A.R.D.S., based on the berlin definition the higher the mortality.(6).

With regards to etiologies, it was noted that A.R.D.S. secondary to trauma was associated with a lower mortality.

(30)

29

Other Studies looking into scores predicting mortality also noted higher APACHE and SAPS scores were associated with a higher mortality. Gong et al in a study on 688 patients in an ICU in Massachusetts(38) found that significant predictors of mortality were increasing age, APACHE score , etiology of trauma (which had a significantly lower mortality) , use of corticosteroids before A.R.D.S. ( Odds ratio of 4.65 for mortality with use) and a pH of <7.22. They also noted that blood transfusions (red blood cells) were associated with an increased odd ratio for mortality of 1.1 per unit transfused.

Another study, done in the All Indian Institute of Medical Sciences (A.I.I.M.S.) Delhi found significant predictors of mortality to be the number of organ failure (Other than pulmonary) , [Hazard ratio 7.65], admission peak pressure [HR 1.13]] and the SAPS-II score ( Simplified Acute Physiology Score) [HR 2.36].(39).

Now with the very large improvements in conservative management, the most common cause of mortality in the initial phase of ARDS is due to the underlying disease that caused the A.R.D.S.

In a 2002 study of 235 patients with ARDS, the most common cause of death was multi organ failure due to sepsis and refractory shock. Some factors suggested in this cohort to predict mortality were te P/F at day 3 and organ dysfunction at day 3.(40)

(31)

30

A cohort done in the west among 207 patients, mortality after day 3 was most commonly due to sepsis, normally from a pulmonary source. Only a small portion (5 of 32 or 16%) of deaths were due to refractory respiratory failure. (41)

Most treatment related measures (Ventilation strategies, prone ventilation, use of paralytics and use of steroids) affecting mortality have already been mention above. Two other factors noted in literature to affect mortality were the fluid balance and timing of intubation. With regards to fluid balance, a secondary analysis of the A.R.D.S. Network cohort noted that a cumulative negative fluid balance by day 4 of A.R.D.S. was associated with a lower hospital mortality (Odds ratio of 0.5, (95% C.I of 0.28-0.89) even after adjusting for confounders. It was also associated with more ventilator free days.

Kangelaris et all followed up 106 patients with A.R.D.S. He noted that in a subset of patients that were not intubated at onset but intubated later (within the first 3 days) after a trail of noninvasive ventilation, the 60 day mortality was higher in the late intubation group (56%) than the early intubation group (36%) and group that did not require intubation (26%), a relationship which was of statistical significance(42) .

Though the significant improvement in the short-term outcomes is encouraging, this has brought to attention a new problem of the overall poor condition of the long-term survivors.

(32)

31

As the mortality of Acute Respiratory Distress Syndrome decreased from approximately 60% to 25% over the past 20 years, the number of survivors after an episode of A.R.D.S. has steadily increased.(43)

Studies have started to now look at the long-term impacts of this disease. As to treat any

problem, the first step is to define the problem, followed by understanding the factors leading to it and then finding treatments to address the pathological process that underlie its development

4.7.3.1 Measurements of long term sequele:

Looking at prior studies and tools at our disposal, the following are some of the possible measuring tools.

4.7.3.1.1 Exertional capacity - The 6 min walk test:

The 6 min walk test is a reliable, validated test for exertional capacity and a predictor of clinical outcomes in respiratory and cardiac disease (44,45).

4.7.3 Long term outcomes:

(33)

32

In this test a subject is asked to walk a corridor of 100mts as quickly as they can but can stop or slow down if they feel fatigue. Its outcomes are the distance walked compared to predicted values based on height and weight and is a good measure of exercise capacity. It also tells us a value known as the distance saturation product, which has been co-related with mortality in patients with idiopathic pulmonary fibrosis, more so than that of distance walked.(46)

Contraindications for the 6-minute walk test are new onset angina or a myocardial infarction in the past 1 month, resting heart rate more than 120 and blood pressure more than 180/100mmHg.

A study by Donald et all, found that on patients with COPD , a value of about 54 mts , corelated with patients ratings of others ability as better or worse to themselves. (47). Recently a

secondary analysis of international studies on A.R.D.S. follow ups showed that when comparing for reported physical health measures, a difference of 20-30 mts could be perceived as

significant for patients.(48)

A study by Michael I. Polkey et all, showed that a difference of 51.8 mts was significant as a predicator for clinical outcomes of death and hospitalization in COPD patients. A decrease in the 6 min walk test of 30 mts or more over one year also had a statistically significant co-relation with clinical outcomes of death and hospitalization .(49).

(34)

33

This was one of the test was used as a measure of physical function in a large cohort by Herridge et all and also was noted to co-relate with the physical aspects of quality of life

questionnaires.(50)

The main limitations of the test are its inability to identify the exact cause of limited exercise capacity. Thus, nutritional problems, musculoskeletal diseases and the contribution of other co- morbid conditions such as cardiac disease to exercise impairment cannot be clearly

distinguished.

The predicted reference Standards for adults age 25-80 has already been suggested for the India population in a study published in lung India (51). However, there was noted to be a significant standard error in the equation for calculating predicted values of 19+/-5%

4.7.3.1.2 Estimation of lung function - Spirometry:

Spirometry has been validated as a method for testing for obstructive and restrictive lung diseases. It has not been significantly used in the context of A.R.D.S. , until recently. (52).

(35)

34

The spirometry provides information on the volume of air that can be exhaled after maximal inspiration and maximal exhaled volume at a specific time period. It can also be used to assess the response to and improvement of the volumes after bronchodilator therapy This allows for assessment of reversibility of an airway obstruction. It has been used to classify the severity of Chronic Obstructive Pulmonary Disease.

The spirometry can be used to asses for an obstructive airway defect, the reversibility of airway obstruction and suggest a possible restrictive airway defect. An FEV1/FVC <0/.7 was suggestive of an obstructive defect as per the Gold guidelines. However, the <5% the lower limit of normal was suggested as a better measure as the value of 0.7 may result in false negatives in younger adults and false positive in older adults(53). A possible restrictive disease is suggested by a low FVC less than 5% the LLN, though the less accurate value of 80% of predicted has been used in some studies. In the setting of an obstructive disease, a low FVC is most likely secondary air trapping due to the severe obstruction but can rarely be due to a mixed disease. Sometimes air trapping, which reduces the FVC can also normalize the FEV1, showing a pseudo restrictive pattern in a severe obstructive disease. For diagnosis of a restrictive disease and differentiating a pseudo restrictive from true restrictive disease, the total lung volumes are needed. The Maximal mid expiratory flow was parameter that was previously used to suggest a small airway disease when values were less than 60%. However, it is noted to have poor sensitivity and specificity for the same.(54)

(36)

35

The prevalence of various spirometry abnormalities in the normal population are 7-11% for the restrictive pattern based on studies in Korea and US(55,56) and 3-20% for obstructive patterns in India. The large variation is due to the difference in states, biomass exposure and smoking(57)

Various studies on pulmonary testing in A.R.D.S. have yielded conflicting results. A study by Schelling et all which had followed up of 50 survivors of A.R.D.S, noted that more the number of impaired components a patient had on pulmonary function tests ( from

FEV1,FVC,FEV1/FVC and DLCO, along with capillary O2 during testing), the more likely they were to report major limitations in their physical and mental quality of life scores in the SF 36 quality of life questionnaire.

Herridge et all in their one-year outcomes of A.R.D.S. found that most of the pulmonary

function tests had normalized by the end of the first year. Only a mild decrease in carbon dioxide diffusion was seen to persist with a gradual improvement of about 9 % in the percentage

predicted of normal at 1 year, from 63% to 72%/. The spirometry at 3 months initially showed a mild restrictive pattern, possibly due to a muscle weakness and persistent lung fibrosis.

However, by the 6 months follow up, FEV1 and FVC had reached nearly 80% of normal in most of the study patients. In fact only a small proportion had persistent pulmonary morbidity after 1 year.(50).

(37)

36

4.7.3.1.3 Quality of life questionnaires:

SF-36 quality of life questionnaire by Rand

The SF-36 questionnaire was a quality of life assessment tool designed by the Rand corporation to assess the quality of life in the domains of vitality, physical functioning, bodily pain, general health perceptions, physical role functioning, emotional role

functioning, social role functioning and mental health.It has been compared to, and been shown to be as useful as other health indexes such as the Nottingham health profile. (58).

Its mental health component has been noted to co-relate well with other mental health scales for depression and PTSD. (59)

The questionnaire is a self-administered form with 36 questions which are later scored to give an 8-domain outcome as mentioned above. It appears to be consistent and

reproduceable.(60)

In the study by Herridge et all, there were significant differences in all domains except emotional role for long term survivors of ARDS, even 5 years after discharge.(50). It was however noted that the physical scores improved dramatically over the first year after illness.

(38)

37

Another finding noted in the study was that the improvements in scores for the physical function and physical domain in SF 36 paralleled the improvements in 6-minute walk test during follow ups.

St. George Respiratory questionnaire

The St. Georges respiratory questionnaire has been used in the monitoring and treatment of respiratory diseases such as asthma, COPD and bronchiectasis. It has been validated in

bronchiectasis(61) . It has been shown to be consistent and representative of the current clinical status of the patients and also co-related well with other functional testing such as the 6 minute walk tests in patients with respiratory disease (62).

It has also been used to judge the efficacy of treatment in asthma and COPD patients, with a score change of 4 to 8 suggesting a mild improvement and more than 12 a significant improvement / very efficacious treatment(53).

In ARDS, the St. George questionnaire was used in a prospective cohort study of long term outcomes of ARDS by Herridge et all and showed that patients had significant morbidity affecting all domains described by the questionnaire during follow up. (63)

4.7.3.2 What data do we have on long term outcomes currently?

(39)

38

One of the most informative studies regarding the long-term outcomes of ARDS was a large cohort study by Herridge et all, first published in NEJM in 2003. The investigators in this study had taken a group of 109 patients with acute respiratory distress syndrome and followed them up at 3,6 and 12 months after discharge. This was done in Toronto between May 1998-2002. They evaluated the patients using a self-administered quality of life questionnaire and pulmonary function tests (the spirometry and 6-minute walk test) at each review. They found persistence of mild to moderate restrictive patterns of pulmonary disease on spirometry with a decreased diffusion capacity for carbon monoxide. There was also initially a decrease in the forced vital capacity and forced expiratory volume at one minute which rapidly improved over 3-6 months.

Chest radiographs were also done and found to be normal in 80 % of all cases. Most patients undergoing the 6-minute walk test, a measure of exertional capacity, could only achieve 60-70%

of the predicted normal distance walked at 6 months and even 1 year after the illness.

The presumed basis given by the authors for this functional limitation was muscle weakness.

This received some support from a possible restrictive pattern noted in spirometry with no correlation with imaging fibrosis and impairment. The main determinant found, for a poor 3- month 6 min walk test, was the use of systemic corticosteroids during admission. At 6 months however the effect of steroids becomes blunted and there is patients are noted to have more significant impairment if they had a longer duration of stay in hospital or a slower rate of recovery (based on the multiple organ dysfunction and lung injury scores used here).

(40)

39

Dr. Herridge later followed all the patients in the cohort for a further 4 more years. The results showed a persistent impairment in the 6-minute walk test, which though better than that of the first year, never reached predicted values expected for a normal population.

Regarding return to work (50) only 16% and 32 % at 3 and 6 months reported being able to return to work. At 6 months, of those who had returned to work, nearly 88% had returned to original work. An extension of the study showed that by 1 year 48% had returned to work and 77% by 5 years(4). In India, where loss of occupation can cause a devastating loss of income to a family, and the urgency to return to work is far greater, we expect the return to work to be

earlier in our population , though possibly at the cost of a lower quality of life. There has been no study available to date studying this. The long-term financial burdens of this disease, in terms of medications, days of wages lost, and direct costs is also largely unknown.

In terms of subjective quality of life, ARDS patients had rated themselves consistently lower in almost all aspects (barring emotional role), though they did show some improvement with time.

The physical and functional disability, as mentioned before, was noted to co-relate with the 6 min walk test. (50) The various predictors of worse quality of life included longer duration of ventilation, concomitant illnesses and time to resolution of organ dysfunctions (64). Surprisingly severity of A.R.D.S. was not a significant predictor of long-term mortality or morbidity.

(41)

40

Another early study on the long-term morbidity of survivors, a prospective cohort by Dr.

Hopkins et all also showed mild cognitive impairment, depression and anxiety even 2 years after the illness. The mental health impairments persisted even up to 2 years.(65)

Figure 2: Mortality, 6-Minute Walk Distance, and Quality of Life scores for 5 Years after Discharge in the cohort by Herridge et al.

(42)

41

(Figure taken from Herridge et all, Functional Disability 5 Years after Acute Respiratory Distress Syndrome, NEJM, April 7 2019(4))

The predominant impairment in the quality of life questionnaires was noted to be in the Mental Health, vitality, social function, physical function and general Health scores. Though there were improvements, particularly in the physical function score form baseline.

Figure 3: Median scores in the SF 36 quality of life questionnaires in the cohort by Herridge et al.

(Figure taken from Herridge et all, Functional Disability 5 Years after Acute Respiratory Distress Syndrome, NEJM, April 7, 2019 (4))

With regards to long term mortality in ARDS, Chen yu Wang et all, did a prospective cohort study of 646 patients with A.R.D.S in Taiwan. The study found a 1-year mortality of 41%. This was noted to be higher than the in-hospital mortality of 21% in this study. Surprisingly there was no relation of the initial severity of A.R.D.S. and one-year mortality. Rather co-morbidities, length of hospital stay and etiology were strong predictors of mortality. (66)However these findings may need to be interpreted with caution as in this study it was noted that a significant

(43)

42

group of patients had a malignancy and it was the primary cause of mortality in one year. The scenario is significantly different from the Indian setting where infections are the main cause of A.R.D.S.

In a retrospective cohort study by Khandelwal et all in 2014 , comprising of 428 patients, it was found that long term survival of patients who had previously been admitted with A.R.D.S. was 85% at 3. (67).

As noted, only 20 % of patients had abnormalities in chest radiography at one year in the follow up by Herridge et al. However, when CT scans were used to check for abnormalities, even 5 years later, there were anomalies in nearly 70% of patient who had A.R.D.S. There was no significant relationship between the radiological abnormalities and ventilatory , exertional and subjective wellbeing of the patients was found. (68)

Was this effect being solely due to the critical illness or did Acute respiratory Distress Syndrome itself play a role in the physical dysfunction. A study by Davidson TA et al matched 73 pairs of ARDS with matched critically ill controls (based on co-morbid illness and APACHE scores) and assessed them for their long-term outcomes based on the prior tested St George Respiratory Questionnaire and F 36 questionnaire. What they noted was that the quality of life for patients

(44)

43

who had Acute Respiratory Distress Syndrome was significantly lower than critically matched controls , suggesting it is not just the critical illness causing the morbidity among survivors. (63)

4.7.3.3 PREDICTORS OF OUTCOME:

The various outcomes can be divided into cognitive, psychiatric, physical function, mortality and return to work. The possible predictive factors are as below.

4.7.3.3.1 With regards to mortality

Co-morbidities and living independently without assistance were significant predictors of mortality after A.R.D.S. Severity of ARDS was not an important predictor for long term outcomes(66). This was as noted by the multi centric prospective cohort study by Wang et all, looking at factors affecting the one-year mortality.

Another Longitudinal prospective cohort of ARDS survivors (156 in toto) noted a significant association of muscle weakness at discharge and mortality at 5 years. Patients who had residual weakness were noted to have a 3 times higher odds of death than those without.(69)

(45)

44

4.7.3.3.2 Cognitive Impairment and psychiatric symptoms-

As noted above, A.R.D.S. survivors had a persistent cognitive impairment well after the resolution of the illness. The prevalence was almost 40% in observational studies(65,70).

The exact pathophysiology is not known, though a lower central venous pressure due to a conservative fluid strategy along with mild hypoxemia were suggested to be contributary.

When looking into studies on hypoxemia in A.R.D.S, in a short term, as per previous studies there was no mortality difference between a lower threshold target and a higher threshold target of oxygen saturation This had allowed for a permissive hypoxia

,translating into lesser ventilatory pressures needed for maintaining oxygenation..

However in the long term, indirect evidence is available for a possible link between hypoxia and cognitive and neurological morbidity.(70) In the ARDS cognitive outcomes study, it was noted that impaired survivors had a lower morning PaO2 than non-impaired survivors (71 vs 86mm Hg). Further studies are needed to confirm this relationship.

4.7.3.3.3 Physical decline

(46)

45

A.R.D.S. resulted in significant impairment to exercise capacity as evidenced by an impaired 6- minute walk tests. In the study by Herridge et all, the 6 minute walk tests were only 66, 67, and 76 percent of the expected value at one , three and five years.(4)

Impaired physical function was observed in as many as 66% of patients in a follow up of 186 patients with A.R.D.S. The main risk factors identified were longer ICU stay and prior

depressive symptoms(71). The rationale for the prolonged ICU stay was greater muscular atrophy and possible neurological dysfunction. The authors of the study proposed that prior depressive symptoms would lead to lesser motivation for rehabilitation, poor eating habits which may impair nutrition needed for recovery, worsen effort during the testing of physical function, increase perceived disability due to disease and also act via neuroendocrine and inflammatory pathways to impair function and hamper recovery.

4.7.3.3.4 Other sequalae

-Complications of intubation and mechanical ventilation such as speech, swallowing impairment and prolonged tracheostomy.

(47)

46

-Delayed return to work, as noted above was a significant issue for all A.R.D.S. patients.

Some factors associated with a longer delay were co-morbidities, longer than 5 days of mechanical ventilation and discharge to another health care facility.(72)

-Impaired quality of life as mentioned above. A 6-month prospective cohort study by Brown et all described multiple variables and their effect on quality of life. They found that 2 lifestyle variables, smoking and obesity had a significant association with lower long-term quality of life. This study noted better outcomes if the individual was

independent for all activities of daily living at baseline.(73)

With regards to ventilation strategies in A.R.D.S as a risk factor for increased morbidity, in a study on 28 survivors, low tidal volume ventilation strategies did not result in any significant impairment with regards to the 6 min walk tests or a respiratory quality of life at follow ups. (74)

(48)

47

4.8 Justification for this study:

A.R.D.S. is a disease with a high prevalence and an increasing number of long-term survivors with high morbidity. Most of the data A.R.D.S. is from a western population.

There is lack of knowledge of the long-term outcome of A.R.D.S. in an Indian population. Much useful information with regards to the risk factors for long term morbidity and overall long-term impact of A.R.D.S. is not yet known. Collecting this data will help us define the scope of the problem in our setting. This will help to identify the patients who are at high risk, identify potentially modifiable risk factors to reduce the risk of morbidity and potential interventions for those with morbidity. We will need to make this first step of identifying the scope of the problem in our population. Also, in a resource-poor country like India, it will be important to understand the economic impact of acute respiratory distress syndrome.

(49)

48

The study protocol was approved by the Institutional Review Board in

[IRB Min No. 11041 dated 04.12.2017] (IRB approval letter in Annexure 11.1.2). The study was funded by the Hospital research fund-Fluid research (Fluid Grant approval Annexure 11.1.3).

The recruitment of participants took place between January 1^st, 2018 and April 30^st 2019 Patients were recruited either during admission or after discharge. Irrespective of the recruitment, all patients were followed up prospectively till October 2019.

This study was a prospective observational cohort study of patients admitted with A.R.D.S. as per the BERLIN Criteria. STROBE checklist was used for designing the study and reporting the outcome (Annexure 11.6)

5 Methods

5.1 Institutional Review Board Approval

5.2 Duration of the study

5.3 Study design

(50)

49

This study was conducted in the Christian Medical College and Hospital, a 2695 bedded University Teaching Institute in south India. It attracts patients from all over India.

However only patients from South India admitted under the general medical units were included in the study. This was for ease of follow up and to minimize loss to follow up.

Patients presenting to the emergency department or admitted into the intensive care unit or in medical wards were screened for recruitment in the study. The following were the inclusion and exclusion criteria.

Inclusion Criteria:

1. All patients satisfying the Berlin definition of A.R.D.S.

2. Patients informed consent is necessary

3. Patients must be from Tamil Nadu, Kerala, Karnataka or Andhra Pradesh 4. Willing for follow up

5. Patients should be older than 16 years of age

All eligible patients were recruited consecutively during the study period.

Exclusion criteria:

-Patients not consenting for follow up.

5.4 Setting

(51)

50

We used the BERLIN CRITERIA for diagnosis of A.R.D.S. Patients are classified as acute respiratory distress syndrome only if all the criteria are met.

Criteria Rationale

Onset within 7 days Observational studies showed A.R.D.S. develops within 72 hours- To 1 week

Bilateral opacities Opacities should not be fully explained by any other single cause.

PaO2/FiO2 ratio < 300 (on a minimum PEEP of 5)

To standardize P/F ratios. Also, oxygen delivery is more reliable with Noninvasive and invasive ventilation systems.

All patients who fulfilled the inclusion criteria were provided with the study Information sheet (Annexure 11.4). Patients were given the opportunity to ask questions, and if willing for participation, an informed consent was taken in the language patients were well versed in. Consent and information sheets were available in regional languages.

mention how you took consent for those recruited after discharge.

5.5 Diagnostic Criteria

5.6 Consent for participation

(52)

51

a. To determine the 6 months outcomes of patients with A.R.D.S

a. To determine the mortality of patients with A.R.D.S., 6 months after discharge b. To calculate the percentage of patients with A.R.D.S. who return to work in the study period.

c. To determine the direct medical costs (hospital bill including medicines) as well as the indirect costs (wages lost during hospitalization as well as time to return to work)

d. To assess the quality of life reported by the patients after an episode of A.R.D.S 6 months after discharge.

e. To identify risk factors that may predict mortality and morbidity.

f. To assess the average distance walked on a 6-minute walk test and outcomes of spirometry in survivors of A.R.D.S., 6 months after illness

5.7 Outcome

5.7.1 Primary Outcome:

5.7.2 Secondary Outcomes:

(53)

52

a. Co-morbidities like diabetes, hypertension, dyslipidemia, smoking, obesity b. Pre-existing cardiac or respiratory disease

c. Family involvement and support for patients d. New illness prior to review

We considered potential sources of Bias prior to initiating the study. We addressed the biases as below.

1) Selection Bias: Since we have taken all consecutive cases, selection bias was limited.

However non -respondent bias was present, with very sick patients or well patients not following up.

2) Information Bias: Information was gathered the same way for all patients with ARDS, irrespective of their severity to limit this bias. However, recall Bias was possible in our study since the prior health noted by the SF36 expand to explain this was a recall of the state just after illness.

5.8 Confounders

5.9 Bias

(54)

53

3) Confounding: We have attempted to include all possible variables that we believe will affect the long-term outcome. This allows for stratification to be done in the analysis.

Unfortunately, we were not able to accurately describe the level of support for disabilities provided after hospital care. Thus, we could only indirectly asses it with the

socioeconomic score.

4) Time related Biases: As some patients were a significant distance away from hospital, other issues such as the need for an accompanying family member, difficulty for further leave from work and patient hesitancy in coming back to a hospital were all factors involved in our follow up. Thus, some patients followed up at dates much later than the planned follow up.

The sample size was calculated using the reduction in walking distance at the end of 6 months. A difference of 50m with a standard deviation of 5m was considered significant.

If we consider a precision of 1m, we need a sample of 100 subjects based on the prior cohort by Herridge et all (50). And if we assume a 50% drop in follow-up, we need 50 more patients, resulting in around 150 (100+50=150) subjects. The sample size calculated for 1m precision and 95% confidence interval and the following formula was used

Formula:

5.10 Sample size calculation

(55)

54

N = 4*sd*sd/d*d

sd= standard deviation, d=precision

The data was collected from the patient or patient’s relative if the patient was unable to furnish the necessary information by direct and telephonic interview using the standard Clinical Research Form (CRF) designed for this study (Annexure 11.3) by the principal investigator. Some of the required data required was collected from the electronic health records of the hospital (Clinical Workstation, Version; Christian Medical College) Department of Computerized Hospital Information Processing Services).

Baseline demographics data included age, gender, hospital number, source of admission, date of admission and discharge, duration of hospitalization, education, occupation, dependency for activities of daily living, income and socioeconomic class- Modified Kuppuswamy socioeconomic scale (Annexure 11.8). Other patient variables including co-morbidities, use of alcohol, smoking and use of tobacco were also collected. Disease variables included etiology of the A.R.D.S., severity of the A.R.D.S, associated

hypotension, secondary infections and complication including an acute kidney injury.

Treatment variables included duration of invasive and noninvasive ventilation, use and

5.11 Data sources and measurement

5.12 Parameters

(56)

55

total dose of steroids, use of paralytics, prone ventilation and vasoactive medications.

Need for interventions such as dialysis, transfusions and their number needed, and tracheostomy were also recorded. The total cost of medical treatment, including medications was recorded.

At review, data on the patients 6-minute walk test and its comparison to predicted models, their spirometry function, a SF 36 quality of life questionnaire and St. George Respiratory questionnaire were all collected. Additional outcomes of loss of wages, duration to return to work and whether they returned to their original work or not was also collected at review.

Data entry was done by the principal investigator from the CRF in Epidata version 3.1.

This data was then exported to SPSS for windows version 17, IBM Corporation. All statistical analysis was done by Mrs. Gowri (Lecturer, Department of Biostatistics, Christian Medical College, Vellore).

Data was entered using epidata software. Data was cleaned by excluding outliers and other data entry mistakes by plotting histogram and boxplot.

All baseline data that were categorical were described using numbers and percentages.

Continuous data were described using mean and standard deviation.

5.13 Data analysis and statistical methods

(57)

56

Categorical variables were summarized using counts and percentages. Quantitative variables were summarized using Mean (SD) for normally distributed variables Median (IQR) for skewed variables.

Descriptive statistics were done for the continuous variables and reported as mean ± SD.

Frequency with percentages were calculated for categorical variables. The Chi square test was used to asses for association and strength of association in the bivariate analysis.

Cross tabulation was done between categories of severity of A.R.D.S. and the outcome variable.

(58)

57

6.1 Strobe Diagram

Figure 4 Strobe Diagram

6 Results

(59)

58

6.2 Baseline characteristics:

Table 2: Baseline characteristics

Characteristic No. (%)

Age (mean +/- SD) 47.39 +/- 14.99

Gender (n= 109) Male

Female

55 (50.5) 54 (49.5) State of origin (n=109)

Tamil Nadu Andhra Pradesh Karnataka Kerala

80 (73.4) 29 (26.6) 0 (0) 0 (0) Marital status (n=109)

Married Unmarried Divorced Widowed

92 (84) 11 (10) 2 (1.82) 4 (3.64) Dependency (n=109)

Independent for activities of daily living Dependent on relatives

106 (97.2%) 3 (2.8)

(60)

59

Needing professional care 0 (0)

Education of head (n= 108) Illiterate

Primary School certificate Middle

High

Intermediate or postgraduate Graduate

Postgraduate

4 (3.7) 8 (7.4) 19 (17.6) 24 (22.2) 32 (29.6) 17 (15.7) 4 (3.7) Occupation of head (n= 106)

Unemployed Unskilled worker Semi-skilled worker Skilled worker

Clerical / Shop-owner / farmer Semi-professional

Professional

3 (2.8) 11 (10.3) 14 (13.1) 17 (16) 32 (29.9) 17 (15.9) 12 (11.2)

Family income per month (n= 106) (Less than ₹ 2181)

(₹ 2181 – ₹ 6477)

78 (73.6) 21 (19.8)

(61)

60

(₹ 6478 - ₹ 10795) (₹ 10796 - ₹ 16193) (₹ 16194 - ₹ 21591) (₹ 21592 - ₹ 43184) (Above ₹ 43184)

27 (25.5) 14 (13.2) 13 (12.3) 15 (14.1) 9 (8.5)

Kuppuswamy Class (n= 106) Upper

Upper Middle Lower Middle Upper Lower Lower

7 (6.6) 26 (24.5) 43 (40.5) 29 (27.3) 1 (0.9)

Salary of earning member (n=108)

Median (25^th – 75^th quartile)

₹ 10,000 (6000-20000)

Anthropometric data (n=81)

Height in cm (mean +/- SD)

Weight in kg (mean +/- SD)

160.8 +/- 8.8

65.6 +/- 16.3

regulations for MD General Medicine examination of the Tamil Nadu

To Study the long-term outcomes of patients with Acute Respiratory Distress Syndrome admitted to general medical wards and intensive care units in a tertiary care hospital in South India (L.A.R.S. Study)

A dissertation submitted in partial fulfilment of the rules and