Keywords
dengue - thrombocytopenia - homeopathy -
Crotalus horridus
Introduction
Dengue is the most common mosquito-borne viral disease in the world and around 50
million dengue infections occur each year. In the South-East Asian region, the case
fatality rate is 1%, but in Myanmar, India, and Indonesia it ranges from 3% to 5%.[1] Dengue is associated with significant morbidity and mortality. It is endemic in
many parts of India including its capital, New Delhi, which is an important metropolitan
hub for major financial and socio-cultural exchange, offering challenging threats
to current public health infrastructure. In the recent past, an upsurge of dengue
cases in New Delhi posed a significant menace to the existing dengue control policies.[2] The reported incidence of dengue, as per the National Vector-Borne Disease Control
Program of India during 2015, was 15,867, with 60 deaths.[3] However, the actual economic and disease burden of dengue in India is substantially
more than official reported cases.[4]
Dengue has several hematological manifestations, including thrombocytopenia and increased
bleeding risk. Thrombocytopenia is reported in 79% to 100% of patients with dengue.[5]
[6]
[7] The proportion of dengue patients receiving platelet transfusion ranges from 7%
to 50%, as reported in studies from Trinidad and Tobago, India, Taiwan, and Singapore.[8] Platelet transfusion, though administered to 22% to 50% of adults in various settings,
could be inappropriate in around 23% of patients.[9] In the absence of major bleeding in dengue with thrombocytopenia, prophylactic platelet
transfusion, with an intention to prevent hemorrhagic complications, is questionable.[10] A recent global survey of physicians indicated the heterogeneity of approaches to
the use of platelet prophylaxis in dengue patients and variations in clinical practice,
with a lack of evidence in this area.[11] This wide range reflects varying local practices and general lack of consensus with
regard to the management of thrombocytopenia in dengue.[6] Thrombocytopenia, one of the important diagnostic criteria for dengue, is being
proposed as an important predictive marker for complications, as studies have shown
that lower platelet count correlates with higher complication rate.[12] Several convergent studies have shown that the death rate in dengue fever is six-fold
higher in individuals whose platelet count is below 50,000/µL.[13]
With no specific treatment available, current recommended treatment is largely supportive,
with careful fluid replacement. There are limited studies in the literature assessing
benefits and risks of different modalities for managing thrombocytopenia in dengue
infection.[14] Presently, there is no effective, commercially available, treatment for dengue.[15] Homeopathic medicine has been advocated for therapeutic use in dengue since the
19th century.[16] The homeopathic medical repertory lists 16 medicines for dengue.[17] Jacobs et al[18] conducted a pilot study with a homeopathic combination for dengue fever symptoms:
the results were inconclusive. Saeed-ul-Hassan et al[19] conducted a comparative study to assess the effect of a different homeopathic combination
on platelet count and found encouraging results, showing improvement in platelet count.
Further, Mahesh et al[20] reported homeopathic treatment of 10 dengue fever cases with positive outcome.
During the dengue outbreak of 2015 in Delhi, patients at the Dr. Hedgewar Aarogya
Sansthan in-patient department (IPD) were given homeopathic treatment along with usual
care (UC) or UC alone. This article reports and evaluates those findings.
Methods
Study Design
This was a comparative cohort study of integrative medical practice in a routine IPD
setting during an epidemic outbreak of dengue. Only patients who gave informed consent
were included in the study. Among them, all patients were offered to receive homeopathy
in addition to UC: those who preferred add-on homeopathy received homeopathic medicines
along with usual care (H+UC); the remaining patients received UC only. Two physicians
(DN and VC), each with more than 10 years of professional experience, were responsible
for prescription of medicines. Four homeopathic physicians (post-graduate qualification),
each with 5 years of experience, meticulously recorded clinical and laboratory data
on a pre-designed case report form. Though these data were collected from routine
practice, ethical clearance was obtained from the Institutional Ethics Committee of
the Central Council for Research in Homoeopathy, New Delhi.
Study Setting
The study was conducted in the IPD of Dr Hedgewar Aarogya Sansthan, an allopathic
and homeopathic hospital in Delhi, during the dengue epidemic outbreak from September
to December 2015. The homeopathic physicians approached the diagnosed/probable dengue
patients admitted to the IPD and, after obtaining consent, adjunctive homeopathic
treatment was provided.
Participants
The criteria for inclusion of patients in the study were as per the national standard
guidelines:[21]
-
Patients compatible with clinical presentation of dengue and seropositive (NS1+ and/or
Mac Elisa IgM+), along with thrombocytopenia (‘Confirmed case’).
-
Patients with fever, or recent history of fever (< 10 days), with dengue-like symptoms
with thrombocytopenia, and from locations where dengue outbreak had been declared
(‘Probable case’).
-
Informed consent to participate in the study.
Laboratory Investigations
As per the standard procedure of the hospital, on admission all patients underwent
the basic laboratory investigations such as hemoglobin (g/dL), packed cell volume,
hematocrit, white blood cell, red blood cell, and platelet count. However, platelet
count measurement of all the included patients was performed on a daily basis from
the day of admission until they reached 100,000/mm3 or more.
Intervention
Patients in the H+UC group were given the indicated homeopathic medicine along with
UC. However, with experience gained during treatment of patients, Crotalus horridus was given in cases where indications for an individualized homeopathic medicine were
deficient. Those in the UC group were given only UC. After the patients were admitted
into the IPD and consent obtained, the indicated homeopathic medicine and UC were
given concurrently.
The UC[21] consisted of bed rest and cold sponging to keep temperature below 39°C; anti-pyretics
were orally administered to lower the body temperature, and oral fluid and electrolyte
therapy were provided for patients with excessive sweating or vomiting. Intravenous
fluids were given to patients as per the need of the individual case. Platelet transfusions
were done as per need and availability of a platelet donor.
Data Collection
Clinical data of each patient were recorded systematically on a pre-designed case-recording
format for age, gender, fever, and key warning signs of illness (e.g., hypotension,
intense abdominal pain, and bleeding), duration of fever, length of hospital stay,
platelet count, number of platelet infusions, and homeopathic prescriptions.
Outcome
The primary outcome assessed was change in platelet count during the course of treatment.
Secondary outcomes were number of platelet infusions and time taken in days to reach
100,000 platelets/mm3.
Sample Size and Statistical Analysis
Sample size was not pre-determined. It involved convenience sampling. All statistical
analyses were performed with IBM SPSS v20. Descriptive data/baseline characteristics
were expressed in number (%), mean ± standard deviation. Categorical data, such as
gender, presence of fever, positivity of NS1/IgM/IgG, presence of bleeding, and type
of bleeding, were also compared in both the groups. Missing values of platelet count
were handled with imputation by the linear interpolation method.[22] Standard error of the mean (SEM)[23] was used for statistical inference based on the sampling distribution. Difference
of platelet count from baseline to each follow-up was analyzed by independent t-test. Kaplan–Meier analysis was performed for time taken to reach the 100,000/mm3 platelet count. A p-value less than 0.05 was considered as statistically significant.
Results
A total of 283 patients, with fever or history of recent fever and who were admitted
to the hospital, were studied (H+UC = 138; UC = 145). The flow diagram of patients
in the study is given in [Fig. 1]. Mean age was 26.3 ± 11.2 years and 27.3 ± 13.2 years in the H+UC and UC groups,
respectively. Seventy percent (n = 198) were male patients. All patients had history of fever as clinical presentation
and only 6% patients presented with fever during admission. Other clinical features
were headache, 18.4%, (n = 52); nausea and vomiting, 38.2% (n = 108); painful abdomen, 44.9% (n = 127); and bleeding, 48.4% (n = 137). One hundred and twenty-one (42.8%) patients had positive serology (NS1/IgM/IgG)
to dengue virus. Mean platelet count was 26,123 ± 16,059/mm3 and 27,724 ± 17,925/mm3 in the H+UC and UC groups respectively, and were comparable at baseline. When considering
the cases for which dengue serology was performed, the mean platelet counts of H+UC
and UC were also comparable at baseline (H+UC = 24,481 ± 17,495/mm3, UC = 25,551 ± 12,945/mm3; p = 0.70). The baseline characteristics are reported in [Table 1].
Table 1
Baseline characteristics of patients
|
Variables
|
H+UC (n = 138)
|
UC (n = 145)
|
p-Value
|
|
Age (years)
|
26.3 ± 11.2
|
27.8 ± 13.2
|
0.27
|
|
Male
|
106 (76)
|
92 (64)
|
0.02
|
|
Female
|
33 (24)
|
52 (36)
|
0.02
|
|
Dengue cases
|
|
|
|
|
Suspected
|
86 (62)
|
76 (52)
|
0.09
|
|
Serology positive
|
52 (38)
|
69 (48)
|
0.09
|
|
Duration of illness (days)
|
7.1 ± 2.4
|
6.1 ± 2.7
|
0.04
|
|
Platelet count (cells/mm3)
|
26,123 ± 16,059
|
27,724 ± 17,925
|
0.43
|
|
Platelet count category
|
|
|
|
|
50,000–100,000 (cells/mm3)
|
13 (10)
|
17 (12)
|
0.29
|
|
20,000–50,000 (cells/mm3)
|
46 (33)
|
48 (33)
|
0.19
|
|
Below 20,000 (cells/mm3)
|
79 (57)
|
80 (55)
|
0.23
|
|
Fever
|
|
|
|
|
Presenting with fever
|
4 (3)
|
15 (10)
|
0.01
|
|
History of recent fever
|
134 (97)
|
130 (90)
|
0.01
|
|
Pulse pressure (mmHg)
|
37.8 ± 8.9
|
33.9 ± 7.5
|
0.01
|
|
Blood pressure (mmHg)
|
|
|
|
|
Systolic
|
103.2 ± 16.9
|
98.8 ± 12.2
|
0.03
|
|
Diastolic
|
70.1 ± 9.1
|
66.2 ± 8.8
|
0.08
|
|
Hb (g/dL)
|
14.1 ± 2.6
|
13.6 ± 2.6
|
0.20
|
|
PCV (%)
|
42.3 ± 7.4
|
40.0 ± 7.1
|
0.01
|
|
Hct (%)
|
41.8 ± 8.8
|
40.3 ± 9.2
|
0.33
|
|
WBC (cells/mm3)
|
1971 ± 3514
|
2320 ± 3154
|
0.43
|
|
RBC (million cells/μL)
|
5.23 ± 4.25
|
4.69 ± 1.00
|
0.28
|
Abbreviations: H+UC, homeopathic medicine plus usual care; Hb, hemoglobin; Hct, hematocrit;
PCV, packed cell volume; RBC, red blood cell; SD, standard deviation; WBC, white blood
cell.
Note: Data presented in n (%), mean ± SD.
Fig. 1 Flow of patients in the study.
Platelet Count
Platelet count of all patients under study was checked every day. There was a significant
rise in platelet count on day 1 of follow-up in the H+UC group compared with UC alone
(mean difference = 12,337/mm3; 95% CI, 5,421 to 19,252; p = 0.001). This trend persisted until day 5 (mean difference = 14,809/mm3; 95% CI, 1,615 to 28,004; p = 0.02). Graphical trend lines for the two groups of patients are shown in [Fig. 2].
Fig. 2 Difference (rise) in mean platelet count from baseline at different days. H+UC, homeopathic
medicine plus usual care. Data show mean and SEM.
Sub-group analysis was performed as per bleeding risk in relation to platelet count[24] in the three categories of high risk (< 20,000/mm3), moderate risk (20,000/mm3 to < 50,000/mm3). and mild risk (50,000/mm3 to 100,000/mm3). A significant difference was observed from day 1 (mean difference = 54,718/mm3; 95% CI, 22,063 to 87,373; p = 0.002) up to day 2 (mean difference = 59,400/mm3; 95% CI, 18,184 to 100,615; p = 0.006) in the mild-risk group. Similarly, in the moderate-risk group, a significant
difference was found at day 2 (mean difference = 18,331/mm3; 95% CI, 4,027 to 32,635; p = 0.013) and day 3 (mean difference = 20,821; 95% CI, 4,828 to 36,814; p = 0.011). In the high-risk group, however, the significant difference was found at
day 2 only (mean difference = 17,533/mm3; 95% CI, 6,655 to 28,411; p = 0.002).
Further, analysis of dengue sero-positive cases showed a statistically significantly
greater increase in platelet count on day 2 of follow-up in H+UC compared with UC
alone (mean difference = 8,522/ mm3; 95% CI, 2,286 to 14,759; p = 0.008).
Time to Reach 100,000/mm3 Platelet Count
A Kaplan–Meier curve was drawn to compare the time taken to reach a platelet count
of 100,000/mm3 ([Fig. 3]). We observed that the patients in the H+UC group reached a platelet count of 100,000/mm3 nearly 2 days earlier than UC alone (H+UC: 3.44 ± SE 0.18 days; 95% CI, 3.08 to 3.80;
UC alone: 5.28 ± SE 0.29 days; 95% CI, 4.71 to 5.86; p < 0.001).
Fig. 3 Kaplan–Meier estimates of time required (days) to reach 100,000/mm3 platelet count. H+UC, homeopathic medicine plus usual care.
Platelet Transfusion
A total of 45.2% (n = 128) of patients received platelet transfusion. Among them, 66% (n = 85) belonged to the high-risk category (i.e., platelet count < 20,000/mm3), out of which 42.3% (n = 36) patients were from H+UC and 57.6% (n = 49) patients were from UC. No significant difference was observed for platelet
transfusion across the risk groups. Out of 159 (56%) patients who had platelet count < 20,000/mm3, 27.6% (n = 44) patients received prophylactic transfusion without bleeding symptoms. A graphical
comparison of number of platelet units transfused between the groups is given in [Fig. 4].
Fig. 4 Between-group comparison of number of platelet units transfused. H+UC, homeopathic
medicine plus usual care. Data show no. of patients along with 95% CI.
Medicines Prescribed
Sixteen homeopathic medicines were prescribed to the patients. Of these, 112 patients
(81.2%) received Crotalus horridus. The next most-prescribed medicines were Phosphorus (4 patients, 3%) and Nux vomica, Pulsatilla nigricans, and Pyrogen (3 patients, 2.1%, for each).
Discussion
Our study suggests that thrombocytopenia of patients with fever or history of recent
fever during dengue epidemic can significantly improve with integrative management.
The data reflect routine practice in managing thrombocytopenia due to dengue with
homeopathic medicines along with usual care. With H+UC, the platelet counts were seen
to increase earlier than with UC alone.
Healthcare during an epidemic outbreak is always a challenge, and more so in developing
countries such as India. Much of India lies in the tropics and is therefore susceptible
to outbreaks of various diseases including dengue. Due to its large population, it
is challenging for health service providers to treat thrombocytopenia. Studies from
different parts of India report high percentages of inappropriate transfusions, ranging
from 13% to 71% in dengue epidemics.[25] Further, in India, blood transfusion services are highly fragmented, and only 35%
of blood units are separated into components. Thus, during explosive outbreaks of
dengue when there is sudden surge in demand for platelet products, managing platelet
inventories becomes a challenge as the transfusion facility also caters for oncology
patients who are major consumers of the same platelet products.[26] As regards preventive measures, a tetravalent vaccine is commercially available.[27] However, the drawback of this vaccine is its limited efficacy and potentially severe
side-effects[28] and thus has limited use in clinical practice.[29] In our study, there was no statistical evidence that the number of platelet infusions
was reduced in association with adjuvant homeopathy.
No clear guidelines exist for the management of thrombocytopenia. The practice of
platelet transfusion has been adapted into standard clinical practice in the management
of hospitalized dengue patients.[30] As per the WHO guidelines,[31] platelet transfusion is indicated in dengue patients having hemorrhage with or without
thrombocytopenia and as prophylaxis in those whose platelet count is below 10,000/mm3 in the absence of active bleeding. In this study, in the high-risk platelet category
(<20,000/mm3), transfusion was required in 42.3% of patients (n = 36) after starting homeopathy treatment compared with 57.6% (n = 49) in the UC group; however, the difference was not statistically significant.
There were 56% (n = 159) patients who had platelet count <20,000/mm3, out of which 27.6% (n = 44) patients received prophylactic transfusion without bleeding symptoms, which
is a smaller proportion than reported for an observational study conducted on transfused
and non-transfused dengue patients.[8]
Our results show that the difference in mean rise in daily platelet count from baseline
until 5th day was greater in the H+UC group as compared with UC. A comparative study,
conducted in Pakistan using a homeopathy combination, found a daily rise in platelet
count as compared with standard maintenance therapy. However, the latter study was
conducted on mild-risk patients only.[19] In another comparative study of prophylactic transfusion versus supportive care
alone, no difference in mean daily platelet count was found between the two groups.[8] However, our study, having the same assessment parameter, showed significant difference
in favor of adjunctive homeopathy (p < 0.05).
The addition of homeopathy treatment to usual care seemed to improve daily platelet
count of dengue patients but had equivocal impact on the requirement for platelet
transfusion. These results may be clinically relevant in a platelet crisis during
a dengue epidemic. Our results may form the basis of larger clinical studies to evaluate
the potential of homeopathy in dengue patients. It is thought that the platelet count
correlates with the severity of the infection and may contribute to the risk of developing
hemorrhage, a well-recognized complication of dengue. Because of this, prophylactic
platelet transfusions are used in some settings to reduce the risk of hemorrhage.
There is currently no evidence to support this practice; moreover, platelet transfusions
are costly and sometimes harmful.[8]
Even though there is a poor correlation of bleeding with platelet counts, it is a
major reason for alarm in both patients and physicians during a confirmed epidemic.
Low platelet counts are often the reason for hospitalization and inappropriate platelet
transfusions in many patients, resulting in further increased morbidity.[8]
[32]
The clinical experience acquired while following this cohort allowed us to implement
the specific management protocol of prescribing Crotalus horridus in cases of thrombocytopenia, where clear indications for an individualized homeopathic
medicine were deficient due to use of medications under usual care. Crotalus horridus was successfully used in a yellow fever outbreak in Philadelphia in 1853,[33] which has a very similar pathogenesis to dengue—that is, fever, hemorrhage, and
low platelet count.[34] About 88% of the H+UC group received Crotalus horridus, which is unique to our study. Jacobs et al[18] and Saeed-ul-Hassan et al[19] both used a homeopathic combination of multiple medicines in their studies. In the
combination used by Saeed-ul-Hassan et al,[19] Crotalus horridus was one component among 10 drugs; a significant difference in platelet count was
noticed on the sixth day of treatment. Jacobs et al's[18] study targeted dengue fever rather than dengue hemorrhagic fever, and Crotalus horridus was not a component of the six-medicine combination used by them. Mahesh et al,[20] in a series of 10 dengue fever cases, prescribed homeopathic medicines based on
individual characteristics. An individualized homeopathic prescription requires the
diagnosis of disease as well as of patient, which is not always easy in emergency
situations such as epidemics and resource-poor settings. The effect of Crotalus horridus in management of thrombocytopenia due to dengue can further be validated in rigorously
designed studies. As dengue has different strains of viruses and their virulence also
varies from outbreak to outbreak, Crotalus horridus may have limited effect with change in virus strain and also in virulence.
Our results are likely to be biased by several factors. Our study primarily registered
adults in Delhi, India, and our findings should not be generalized to Indian children
or to patients in other geographical regions. Second, this study collected data from
patients in routine practice; therefore, there was the possibility of not obtaining
all the investigational values for all the patients. Third, serology for dengue was
not performed in 57% of patients. Nevertheless, the predictive power of clinical diagnosis
will be high during an epidemic because of increased background prevalence of the
disease. Moreover, women were under-represented in our study (30%), though the female
gender is a risk factor for bleeding in dengue.[35] Consequently, referral patterns and diagnostic procedures might bias the patient
sample and clinical characteristics. In addition, we lacked serotype information of
dengue viruses, as well as information on immune status, for a substantial number
of patients. There is also the possibility of preference bias, as only patients who
preferred to take adjuvant homeopathic intervention received it along with usual care.
Effect of patient preference varies with setting.[36] Our study was conducted in an IPD setting and the outcome was an objective/quantitative
parameter; therefore, any preference bias may have been minimal.
Conclusion
Our results suggest a positive role of adjuvant homeopathy in thrombocytopenia due
to dengue. Randomized controlled trials may be conducted to add more insight into
this integrative approach.
