Cochrane Ayurveda Schizophrenia

Cochrane Database of Systematic Reviews
Ayurvedic medicine for schizophrenia (Review)
Agarwal V, Abhijnhan A, Raviraj P
Agarwal V, Abhijnhan A, Raviraj P.

Ayurvedic medicine for schizophrenia.
Cochrane Database of Systematic Reviews 2007, Issue 4. Art. No.: CD006867.
DOI: 10.1002/14651858.CD006867.
www.cochranelibrary.com
Ayurvedic medicine for schizophrenia (Review)

Copyright © 2010 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
TABLE OF CONTENTS
HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Analysis 1.1. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 1 Leaving the study
early. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Analysis 1.2. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 2 Mental state: 1. Not
improved. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Analysis 1.3. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 3 Mental state: 2. Unco-
operative or did not comprehend. . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Analysis 1.6. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 6 Behavioiur: Average
score (Fergus Falls, high score = poor). . . . . . . . . . . . . . . . . . . . . . . . . . 23
Analysis 1.7. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 7 Psychological assessment:
1. Critical flicker fusion threshold (simple). . . . . . . . . . . . . . . . . . . . . . . . 23
Analysis 1.9. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 9 Adverse effects. . 24
Analysis 2.1. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 1 Leaving the
study early. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Analysis 2.2. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 2 Mental state:
1. Not improved. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Analysis 2.3. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 3 Mental state:
2. Unco-operative or did not comprehend. . . . . . . . . . . . . . . . . . . . . . . . . 27
Analysis 2.8. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 8 Behavioiur:
Average score (Fergus Falls, high score = poor). . . . . . . . . . . . . . . . . . . . . . . 29
Analysis 2.9. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 9 Psychological
assessment: 1. Critical flicker fusion threshold (simple). . . . . . . . . . . . . . . . . . . . 30
Analysis 2.11. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 11 Adverse
effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Analysis 3.1. Comparison 3 AYURVEDIC TREATMENT versus ANTIPSYCHOTIC (all short term), Outcome 1 Leaving
the study early. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Ayurvedic medicine for schizophrenia (Review) i

[Intervention Review]
Ayurvedic medicine for schizophrenia
Vishesh Agarwal1 , Akhil Abhijnhan2 , Prakash Raviraj3

1 Jaipur, India. 2 Kalyan, Kochi, India. 3 Wakefield, UK
Contact address: Vishesh Agarwal, 46A, Krishna Nagar, Dher Ka Balaji, Jaipur, Rajasthan, 302023, India. vishesh.dr@googlemail.com.
Editorial group: Cochrane Schizophrenia Group.

Publication status and date: Edited (no change to conclusions), published in Issue 3, 2010.
Review content assessed as up-to-date: 18 August 2007.
Citation: Agarwal V, Abhijnhan A, Raviraj P. Ayurvedic medicine for schizophrenia. Cochrane Database of Systematic Reviews 2007,
Issue 4. Art. No.: CD006867. DOI: 10.1002/14651858.CD006867.
ABSTRACT
Background
Ayurvedic medicine has been used to treat mental health problems since1000 BC.
Objectives
To review effects of Ayurvedic medicine or treatments for schizophrenia.
Search methods
We searched the Cochrane Schizophrenia Group Trials Register (March 2007) and AMED (March 2007), inspected references of all
identified studies and contacted the first author of each included study.
Selection criteria
We included all clinical randomised trials comparing Ayurvedic medicine or treatments with placebo, typical or atypical antipsychotic
drugs for schizophrenia and schizophrenia-like psychoses.
Data collection and analysis
We independently extracted data and calculated random effects, relative risk (RR), 95% confidence intervals (CI) and, where appropriate,
numbers needed to treat/harm (NNT/H) on an intention-to-treat basis. For continuous data, we calculated weighted mean differences
(WMD).
Main results
From the three small (total n=250) short included studies, we were unable to extract any data on many broad clinically important
outcomes such as global state, use of services, and satisfaction with treatment. When Ayurvedic herbs were compared with placebo,
about 20% of people left the studies early (n=120, 2 RCTs, RR 0.77 CI 0.37 to 1.62). Mental state ratings were mostly equivocal with
the exception of the brahmyadiyoga group using Ayurvedic assessment (n=68, 1 RCT, RR not improved 0.56 CI 0.36 to 0.88, NNT
4 CI 3 to 12). Behaviour seemed unchanged (n=43, 1 RCT, WMD Fergus Falls Behaviour Rating 1.14 CI -1.63 to 3.91). Nausea
and vomiting were common in the brahmyadiyoga group (n=43, RR 13.13 CI 0.80 to 216.30). When the Ayurvedic herbs were
compared with antipsychotic drugs (chlorpromazine), again, equal numbers left the study early (n=120, 2 RCTs, RR for brahmyadiyoga
0.91 CI 0.42 to 1.97) but people allocated herbs were at greater risk of no improvement in mental state compared to those allocated
chlorpromazine (n=45, RR 1.82 CI 1.11 to 2.98). Again, nausea and vomiting were found with use of brahmyadiyoga (n=45, 1 RCT,
RR 20.45 CI 1.09 to 383.97, NNH 2 CI 2 to 38). Finally, when Ayurvedic treatment, in this case a complex mixture of many herbs, is
compared with chlorpromazine in acutely ill people with schizophrenia, it is equally (~10% attrition, n=36, RR 0.67 CI 0.13 to 3.53),
but skewed data does seem to favour the chlorpromazine group.
Ayurvedic medicine for schizophrenia (Review) 1
Authors’ conclusions
Ayurvedic medication may have some effects for treatment of schizophrenia, but has been evaluated only in a few small pioneering
trials.
PLAIN LANGUAGE SUMMARY
Ayurvedic medicine for schizophrenia
Ayurvedic medicine was developed in India over 3000 years ago and is the oldest medical system to have survived until the present
time. It sees each individual as having a unique mind-body constitution and set of life circumstances. It is similar to traditional
Chinese medicine in believing that matter and energy are the same thing. Treatment in an ayurvedic system is holistic, involving natural
medicine, massage, diet and the regulation of lifestyle. Ayurveda has been used for the treatment of schizophrenia, a serious long-term
mental health condition, since its formulation (c1000 BCE) although nowadays Western-style medication using antipsychotics and
hospital treatment are also used.
This review examines randomised controlled trials which compare aspects of ayurvedic medicine with the use of antipsychotics for
people with schizophrenia. All trials took place in India and were for 12 weeks or less. When the ayurvedic herbs brahmyadiyoga and
tagara were compared to placebo (2 trials) there was no significant difference between the two groups in acceptability of treatment or
overall improvement. The brahmyadiyoga group did, however, show some improvement when assessed ayurvedically (a combination of
assessing aspects of the mind, decision, orientation, memory and habit, and looking for the absence of symptoms of illness). When these
two herbs were compared to groups of people taking the antipsychotic, chlorpromazine, again there was no difference in acceptability
of treatment, but in one of the two trials there was an improvement in mental state in those taking chlorpromazine. There was also a
trial comparing an ayurvedic package (of herbs and other treatment) to chlorpromazine, and although both treatments were acceptable,
the rest of the data were not able to be used. Brahmyadiyoga and tagara tended to have vomiting and nausea as an adverse effect, while
chlorpromazine caused people to be sleepy. It may be possible that ayurvedic treatments could be used as adjuncts to antipsychotic
medication. A new larger trial comparing ayurvedic herb(s) alone, chlorpromazine alone and both together would answer this question.
(Plain language summary prepared for this review by Janey Antoniou of RETHINK, UK www.rethink.org).
BACKGROUND branches: Kayachikitsa (internal medicine); Salya tantra (surgery);

Salakya tantra (ophthalmology and ENT); Kaumarabrhtya (pae-
Developed in India more than 3000 years ago (Subbarayappa diatrics, obstetrics and gynaecology); Agadatantra (toxicology);
2001), Ayurvedic medicine or Ayurveda is one of the oldest Rasayana (geriatrics and nutrition); Vajikarana (sexology); Bhuta
medical systems known. Many other ancient systems of natural vidya (psychiatry and demonology) as described in Ayurvedic clas-
medicine are believed to have their roots in Ayurveda. sics. The two Ayurvedic classics are the Charaka samhita and the
Ayurveda is a Sanskrit word for the ’Knowledge of Life, defining Susruta samhita (Subbarayappa 2001).
the trinity of life as body, mind and spiritual awareness. Knowl-
edge arranged systematically with logic becomes science. It is a Ayurveda recognises the fundamental importance of examination
complete and holistic science of healthy balanced living that views by direct perception (pratyaksa) and inference (anumana). In ad-
each person as an individual, with a unique mind-body constitu- dition, it also accepts verbal or textual knowledge (aptopadesa)
tion and set of life circumstances. Ayurveda and Traditional Chi- and experimentation (yukti). A physician of Ayurveda carries out
nese Medicine are very similar, being based on universal natural a physical examination using his five senses (pancendriya pariksa)
bi-polar concepts that matter and energy are one. Ayurveda aims and elicits patient history (prasna) (Subbarayappa 2001). Therapy
to promote health and cure disease through holistic methods in- consists of three major categories: divine therapy (daiva vyapas-
volving natural medicine, diet, regulated lifestyle (Neddermeyer raya), rational therapy (yukti vyapasraya cikitsa) and psychother-
2006, Lad 1996). The Ayurvedic compendium comprises eight apy (sattvavajaya) (Singh 1994). There are more than 400 drugs
listed in The Ayurvedic Pharmacopoeia of India (API-2006). 3. Placebo or no treatment.
4. Antipsychotic drugs produced by pharmaceutical companies:
Antipsychotic drugs are the mainstay of treatment of schizophrenia
any compound, dose, pattern or means of administration.
today since their development in the 1950s. The role of Ayurveda
in treatment of mental illnesses, however, dates back to the Vedic
period (c. 1000 BC) (Subbarayappa 2001), and it is still being used Types of outcome measures
as sole treatment or in conjunction with antipsychotic medication
1. Leaving the studies early (any reason, adverse events, inefficacy
in both developing and developed countries (Walter 1999). For
of treatment)
example, nearly three quarters of people living in the USA of
2. No clinically important response as defined by the individual
Asian origin used some kind of Complementary and Alternative
studies (e.g. global impression less than much improved or less
medicine in the past 12 months (Hsiao 2006). Also these drugs
than 50% reduction on a rating scale)*
are socially acceptable, better tolerated and less expensive than
3. Global state
conventional modern treatments (Russinova 2002). We know of
3.1 No clinically important change in global state (as defined by
no systematic review of their effects.
individual studies)*
3.2 Relapse (as defined by the individual studies)
4. Mental state (with particular reference to the positive and neg-
OBJECTIVES ative symptoms of schizophrenia)
To review the effects of Ayurvedic medicine compared with an- 4.1 No clinically important change in general mental state score
tipsychotic, placebo or no treatment for people with schizophre- 4.2 Average endpoint general mental state score
nia. 4.3 Average change in general mental state score
4.4 No clinically important change in specific symptoms (positive
symptoms of schizophrenia, negative symptoms of schizophrenia)
4.5 Average endpoint specific symptom score
METHODS
4.6 Average change in specific symptom score
5. General functioning
5.1 No clinically important change in general functioning
Criteria for considering studies for this review 5.2 Average endpoint general functioning score
5.3 Average change in general functioning score
6. Quality of life/satisfaction with treatment*
Types of studies 6.1 No clinically important change in general quality of life
We included all relevant randomised controlled trials. Where a 6.2 Average endpoint general quality of life score
trial was described as ’double-blind’ but it was implied that the 6.3 Average change in general quality of life score
study was randomised, we included these trials in a sensitivity 7. Service use
analysis. If there was no substantive difference within primary 7.1 Number of patients hospitalised*
outcomes (see ’Types of outcome measures’) when these ’implied 8. Adverse effects*
randomisation’ studies were added, then we included these in the 8.1 Number of participants with at least one adverse effect
final analysis. If there was a substantive difference, we only used 8.2 Clinically important specific adverse effects (cardiac effects,
clearly randomised trials and the results of the sensitivity analysis death, movement disorders, prolactin increase and associated ef-
were described in the text. We excluded quasi-randomised studies, fects, weight gain, effects on white blood cell count)
such as those allocated by using alternate days of the week. 8.3 Average endpoint in specific adverse effects
8.4 Average change in specific adverse effects
We divided outcomes into primary (*) and secondary, and into
Types of participants
short-term (up to 12 weeks), medium term (13-52 weeks) and
We included people with schizophrenia, schizophreniform psy- long term (more than one year).
chosis and schizophrenia-like illnesses, diagnosed by any criteria.
Types of interventions Search methods for identification of studies

1. Ayurvedic medicine (plant, animal, mineral or psychological), 1. Electronic searches
not used as an adjunct to other treatments: any dose or combina- 1.2 We searched the Cochrane Schizophrenia Group Trials Regis-
tion. ter (March 2007) using the phrase:
2. Ayurvedic medicine (plant, animal, mineral or psychological), [((*ayurved* or * brahm* or *hindu* or *siddha* or *unmada*) in
used as an adjunct to other treatments: any dose or combination. title, abstract and index fields in REFERENCE) OR ((*ayurved*

or *brahm* or *hindu* or *siddha*) in interventions field in list of those awaiting assessment until further information could
STUDY)] be obtained.
This register is compiled by systematic searches of major databases, 3. Data management
hand searches and conference proceedings (see Group Module). 3.1 Data extraction
1.3 We searched AMED (Allied and Complementary Medicine - We (VA and AA) independently extracted data from selected trials.
1985 to March 2007) (March 2007) using the phrase: When disputes arose, we attempted to resolve these by discussion
[((exp clinical trials/ or exp randomized controlled trials/ or exp with PR. When disputes could still not be resolved we did not
double-blind method/ or exp random allocation/ or randomized enter the data, but added the outcome of the trial to the list of
controlled trial.pt. or clinical trial.pt. or controlled clinical trial.pt. those awaiting assessment.
or clinic$ adj4 trial$).mp. or (random$ adj5 (assign$ or allo- 3.2 Intention to treat analysis
cat$ or assort$)).mp. or (randomi$ adj5 control$ adj5 trial$).mp. We excluded data from outcomes where more than 50% of par-
or (crossover or cross-over).mp. or ((singl$ or doubl$ or trebl$ ticipants in any group were lost to follow up (this does not include
or tripl$) adj (blind$ or mask$)).mp.) and (exp mental disor- the outcome of ’leaving the study early’). In studies with less than
ders/) and (ayrved$.mp. or ayurved$.mp. or ayruvedic.mp. or exp 50% dropout rate, we considered people leaving the study early
ayurvedic medicine/)] to have had the negative outcome, except for the event of adverse
2. Reference lists effects and death.
We inspected references of all identified studies (included and We analysed the impact of including studies with high attrition
excluded) for further relevant trials. rates (25-50%) in a sensitivity analysis. If inclusion of data from
3. Personal contact this group did result in a substantive change in the estimate of
We contacted the first author of each included study for informa- effect, we did not add this data to trials with less attrition, but
tion regarding unpublished trials and extra data on the published presented the data separately.
trials. 4. Data analysis
4.1 Binary data
For binary outcomes we calculated an estimate of the relative risk
Data collection and analysis (RR) (or risk difference (RD) when pooled data included studies
with no event scores in both groups) and its 95% (fixed effect)
1. Selection of studies
confidence intervals (CI). Where possible, we also calculated the
We (VA and AB) independently inspected all reports. We resolved
number needed to treat/harm (NNT/NNH) statistic. Where het-
any disagreement by discussion with (PR). Where there was still
erogeneity was found (see section 5) we investigated the reasons
doubt, we acquired the full article for further inspection. Once
for this.
the full articles were obtained, we (VA and AA) decided whether
4.2 Continuous data
the studies met the review criteria. If disagreement could not be
4.2.1 Skewed data:
resolved by discussion with PR, we sought further information
Continuous data on outcomes in trials relevant to mental health
and added these trials to the list of those awaiting assessment.
issues are often not normally distributed. To avoid the pitfall of
2. Assessment of methodological quality
applying parametric tests to non-parametric data we applied the
We assessed the methodological quality of included studies using
following standards to continuous final value endpoint data before
the criteria described in the Cochrane Handbook (Higgins 2005),
inclusion: (a) standard deviations and means were reported in the
which is based on the degree of allocation concealment. Poor con-
paper or were obtainable from the authors; (b) when a scale started
cealment has been associated with overestimation of treatment ef-
from zero, the standard deviation, when multiplied by two, should
fect (Schulz 1995). Category A includes studies in which alloca-
be less than the mean (otherwise the mean is unlikely to be an
tion has been randomised and concealment is explicit. Category B
appropriate measure of the centre of the distribution - Altman
studies are those which have randomised allocation but in which
1996); In cases with data that are greater than the mean they were
concealment is not explicit. Category C studies are those in which
entered into ’Other data’ table as skewed data. If a scale starts from
allocation has neither been randomised nor concealed. Only trials
a positive value (such as PANSS, which can have values from 30 to
that are stated to be randomised (categories A or B of the hand-
210) the calculation described above in (b) should be modified to
book) will be included in this review. The categories are defined
take the scale starting point into account. In these cases skewness
below:
is present if 2SD>(S-Smin), where S is the mean score and Smin is
A. Low risk of bias (adequate allocation concealment)
the minimum score. We reported non-normally distributed data
B. Moderate risk of bias (some doubt about the results)
(skewed) in the ’other data types’ tables.
C. High risk of bias (inadequate allocation concealment).
For change data (mean change from baseline on a rating scale)
When disputes arose as to which category a trial should be allo-
it is impossible to tell whether data are non-normally distributed
cated, again we attempted resolution by discussion. When this was
(skewed) or not, unless individual patient data are available. Af-
not possible we did not enter the data and added the trial to the

ter consulting the ALLSTAT electronic statistics mailing list, we Firstly, we considered all the included studies within any compari-
presented change data in RevMan graphs to summarise available son to judge for clinical heterogeneity. Then we visually inspected
information. In doing this, we assumed either that data were not graphs to investigate the possibility of statistical heterogeneity. We
skewed or that the analysis could cope with the unknown degree supplemented this by using primarily the I-squared statistic. This
of skew. provides an estimate of the percentage of variability due to hetero-
4.2.2 Final endpoint value versus change data geneity rather than chance alone. Where the I-squared estimate
Where both final endpoint data and change data were available was greater than or equal to 50%, we interpreted this as indicat-
for the same outcome category, we only presented final endpoint ing the presence of considerable levels of heterogeneity (Higgins
data . We acknowledge that by doing this much of the published 2003). Where heterogeneity was present, reasons for this were
change data may be excluded, but argue that endpoint data is more investigated. If it substantially altered the results, data were not
clinically relevant and that if change data were to be presented summated, but presented separately and reasons for heterogeneity
along with endpoint data, it would be given undeserved equal investigated.
prominence. We are contacting authors of studies reporting only 6. Addressing publication bias
change data for endpoint figures. We entered all data from the included studies into a funnel graph
4.2.3 Summary statistic: For continuous outcomes we estimated a (trial effect against trial size) in an attempt to investigate the like-
weighted mean difference (WMD) between groups using a fixed lihood of overt publication bias (Egger 1997).
effects model. Again, if heterogeneity was found (see section 5) we 7. General
investigated reasons for this. Where possible, we entered data in such a way that the area to
4.2.4 Rating scales: A wide range of instruments are available to the left of the line of no effect indicated a favourable outcome for
measure mental health outcomes. These instruments vary in qual- Ayurvedic medicine.
ity and many are not valid, or are ad hoc. Unpublished instruments
are more likely to represent statistically significant findings than
those that have been put into print (Marshall 2000). Therefore,
we included only continuous data from rating scales if the mea- RESULTS
suring instrument had been described in a peer-reviewed journal
and the instrument was either a self report or completed by an
independent rater or relative (not the therapist). Description of studies
4.2.5 Cluster trials
See: Characteristics of included studies; Characteristics of excluded
Studies increasingly employ ’cluster randomisation’ (such as ran-
studies.
domisation by clinician or practice) but analysis and pooling of
For substantive descriptions of the studies please see Included and
clustered data poses problems. Firstly, authors often fail to account
Excluded Studies tables.
for intra class correlation in clustered studies, leading to a ’unit
1. Excluded studies
of analysis’ error (Divine 1992) whereby p values are spuriously
We have excluded four studies in this review. All four were full
low, confidence intervals unduly narrow and statistical significance
journal articles (Farag 2003; Fozdar 1962; Hakim 1964; Weiss
overestimated. This causes type I errors (Bland 1997; Gulliford
1986). Fozdar 1962 and Weiss 1986 were not randomised. Farag
1999).
2003 was randomised but recruited people with sleep onset in-
Where clustering was not accounted for in primary studies, we
somnia and not those suffering from schizophrenia. We have not
presented the data in a table, with a (*) symbol to indicate the
included all the studies included in the chapter on “Review of
presence of a probable unit of analysis error. In subsequent ver-
Previous Studies” from Ramu 1999b as they were all case series.
sions of this review we will seek to contact first authors of studies
2. Awaiting assessment
to obtain intra-class correlation co-efficients of their clustered data
Two studies are currently awaiting assessment. Both reports are
and to adjust for this using accepted methods (Gulliford 1999).
minimal and contain no usable data. Interestingly, none of
Where clustering has been incorporated into the analysis of pri-
these studies were identified by our standard but comprehensive
mary studies, we will also present these data as if from a non-clus-
database search. Ramu 1985 was picked up from our extra search
ter randomised study, but adjusted for the clustering effect.
conducted on AMED. Our web search led us to the CCRAS (Cen-
We have sought statistical advice and have been advised that the
tral Council for Research in Ayurveda and Siddha) website and
binary data as presented in a report should be divided by a ’design
correspondence with them gave us the book Ayurvedic Manage-
effect’. This is calculated using the mean number of participants
ment of Unmada (Schizophrenia). Roy 1964 was obtained from
per cluster (m) and the intraclass correlation co-efficient (ICC)
this book. We have contacted the relevant authorities at CCRAS
[Design effect = 1+(m-1)*ICC] (Donner 2002). If the ICC was
to acquire more information on Roy 1964.
not reported it was assumed to be 0.1 (Ukoumunne 1999).
3. Ongoing studies
5. Test for heterogeneity
We are not aware of any ongoing studies.

4. Included studies compared the Ayurvedic medication with chlorpromazine (200-
We have included three studies in this review. Two, Mahal 1976 450 mg/day). Mahal 1976 and Ramu 1999 also used placebo as a
and Ramu 1992, were identified through our electronic search of comparator. All medications were administered orally.
the Cochrane Schizophrenia Group Trials Register. Ramu 1999 4.6 Outcomes
was obtained from a book sent to us by the Central Council for Mahal 1976 compared brahmyadiyoga and tagara with placebo
Research in Ayurveda and Siddha, in New Delhi, India. Only and reported usable data on mental state and leaving the study
Mahal 1976 was described as randomised. Mahal 1976 and Ramu early. Ramu 1999 also reported usable data on mental state and
1999 were said to be double blinded. leaving the study early for the brahmyadiyoga versus placebo com-
4.1 Length of trials parison. Ramu 1992 did not compare Ayurvedic treatment to
Schizophrenia is a lifelong illness that affects young people. All the placebo. Some of the data provided by Mahal 1976 and Ramu
included studies reported data on short-term follow-up (up to 12 1999 were very skewed and we have reported these in other data
weeks). tables.
4.2 Participants Mahal 1976 and Ramu 1999 compared Ayurvedic herbs (brah-
In Ramu 1992 participants had been diagnosed with schizophre- myadiyoga) with antipsychotic chlorpromazine. Mahal 1976 also
nia using ICD-9 classification. While one study used the diagnos- compared tagara (another Ayurvedic herb) with chlorpromazine.
tic criteria for schizophrenia in accordance with the NIMHANS Both studies recorded usable data on mental state and number
collaborative study (Mahal 1976), another, Ramu 1999, stated of participants leaving the study early. Some of the data provided
hat they used the WHO glossary of mental disorders to diagnose by Mahal 1976 and Ramu 1999 were very skewed and we have
schizophrenia. Participants from all the three studies had been reported these in other data tables.
diagnosed, using the Ayurvedic classification, as unmada. Ramu Only Ramu 1992 compared Ayurvedic treatment (complex com-
1992 employed the Charaka Ayurvedic definition while another bination of a number of medications) with chlorpromazine and
used the Laksana Ayurvedic definition for schizophrenia. Ramu reported data on mental state and leaving the study early. Most of
1992 did not report the sex of participants. Though the male to fe- the data reported on mental state were very skewed and we have
male ratio was uneven in the other two studies, the total number of reported these in other data tables.
participants on pooling both the studies did not vary significantly 4.6.1 Outcome scales: details of the only scales that provided usable
with respect to sex. The mean age of participants from all three data are shown below. Reasons for exclusions of data are given
studies was about 28 years. Ramu 1992 and Ramu 1999 recruited under ’Outcomes’ in the ’Included studies’ table.
people suffering from schizophrenia over a period of two to ten 4.6.1.1 Psychotic Symptom Rating Scale - PSRS (Rockland 1965)
years, while Mahal 1976 only recruited people with less than a two This rating scale provides a quantified profile of the reflection of
year history of schizophrenia. We could not elicit any exclusion clinical judgement, impressions of degree and nature of psychotic
criteria from Ramu 1999 while the other two studies excluded symptomatology. The normality of behaviour, effect or cognition
stupor, epilepsy and learning disability. No trial gave information is indicated by the zero score, where as the psychopathology in
about the severity of illness, although Ramu 1992 mentioned that these areas increases, the score becomes positive or negative. The
participants had active psychotic symptoms for a minimum dura- whole scale total scores are used as quantitative measures of psy-
tion of one month. chosis to compare patient to patient and to compare the patient
4.3 Setting over time. The sub totals in each category indicate whether the
Ramu 1999 was described as taking place in an inpatient depart- major portion is in the behaviour, affect or cognition. Mahal 1976
ment while Mahal 1976 and Ramu 1992 were described as taking and Ramu 1999 used of this rating scale.
place in hospital settings. 4.6.1.2 Brief Psychiatric Rating Scale (Overall 1962)
4.4 Study size This is a brief rating scale used to assess the severity of a range of
All three studies are small. Mahal 1976 with 136 participants was psychiatric symptoms, including psychotic symptoms. The orig-
the largest of the studies. Ramu 1999 included 78 people while inal scale has 16 items (although a revised 18-item scale is com-
Ramu 1992 recruited only 36. monly used). Each item is defined on a seven-point scale varying
4.5 Interventions from ’not present’ to ’extremely severe’, scoring from 0-6 or 1-
The main Ayurvedic herb used in all three studies is brah- 7. Total scores can range from 0-126, with high scores indicating
myadiyoga. When brahmyadiyoga was used on its own (Mahal more severe symptoms. Only Ramu 1992 used this scale to assess
1976; Ramu 1999), it was administered at a mean dose of 12 the changes in mental state.
mg/day at a range of 8-16 mg/day in four times a day schedules. 4.6.1.3 Ayurvedic Assessment
In Ramu 1992, where a complex Ayurvedic treatment was given, Ayurvedic assessment is comprised of two separate assessments.
brahmayadiyoga was one of the components and was administered The first assessment involves the examination of mental aspects:
at a dose of 500 mg. The other Ayurvedic herb administered was Manas (mind), Buddhi (decision), Sanjnana(orientation and re-
tagara (Mahal 1976) at a dosage of 8-12 mg/day. All three studies sponsiveness), Smrti (memory), Bhakti (desire), Sila (habit), Cesta

(psychomotor activity) and Acara (conduct); in a proforma struc- mentioned the method of allocation concealment. This conceal-
tured for this purpose. In the second assessment, the presence or ment of allocation has repeatedly been shown to be of key im-
absence of Laksana (signs and symptoms) of Dosaga Unmada after portance in excluding selection biases (Jüni 2001), therefore, all
the treatment period is noted. The guidelines for completing this trials have been assigned to category B (moderate risk of bias, see
proforma can be obtained from Ramu 1999c. Methods).
4.6.1.4 Fergus Falls Behaviour rating scale (Meyer 1953) 2. Blindness
This is an early ward behaviour scale which has been used with a Two studies were described as being double-blind (Mahal 1976;
broad range of hospitalised patients. It provides a single summed Ramu 1999). One, Ramu 1992, was described as rater-blinded.
score, and is a quick, convenient, fairly objective method of rating None of the studies described the process of blinding. Testing of
behaviour, with which even inexperienced raters obtained greater blinding was not reported in any of the three studies.
than 90% agreement. The scale is of the semi-checklist type. Sever- 3. Loss to follow up
ity judgements are not necessary. Instead, for each statement, the None of the studies attempted to follow up those who had left
rater has to indicate whether the item does or does not apply to the study early. While Mahal 1976 reported “escape and leaving
the condition of the patient. Detailed rating instructions are pro- against medical advice” as the reason for 28 people leaving early,
vided which guarantee high inter-rater reliability. One of the 11 Ramu 1999 did not give any specific reason why 13 people discon-
categories, ’responds to electroshock or insulin therapy’, reflects tinued. Ramu 1992 did not give specific reasons for two people
the scale’s age and should be rephrased to ’response to therapy’. leaving early, although this study did account for the remaining
This would make the scale more adequate and would probably three who discontinued. None of the studies had dropout rates
not affect its reliability or validity. The items cover most of the greater than 40%.
areas considered to be essential components of ward behaviour 4. Data reporting
scales. In spite of its simplicity for use and good coverage of ward Overall most of the data we found could be used. Findings which
behaviour, this scale seems to have been replaced by those more are presented as graphs, in percentiles or just reported as p-val-
recently developed. The scale requires seven to ten minutes for ues are often of little use to a reviewer. Ramu 1992 failed to pro-
completion. vide standard deviations along with the mean values for the out-
4.6.2 Redundant data comes of a psychological assessment conducted to observe cogni-
Enormous efforts are usually invested in studies rating and record- tive changes. Mahal 1976 reported on the spiral after effect out-
ing data that are then reported in such a way as to render them comes of just less than 40% of the trial participants. These partic-
useless for reviews such as this. For example, in Ramu 1992 the ular outcomes from these two trials could not be used for this re-
trialists compared Ayurvedic treatment with chlorpromazine. The view. We are seeking further data regarding these outcomes. Ramu
outcome of change in mental state using Ayurvedic assessment was 1999 had presented the results of adverse effects experienced by
reported in an obscure manner in that they reported the means the participants in a slightly awkward and complicated fashion.
without the standard deviations. Such results are of no value to However we have been able to extract these data.
us. We have tried to contact the trialists to obtain the standard
deviations. Similarly in Mahal 1976 the data provided on spiral
after effect had attrition rates >40%, that the data was considered Effects of interventions
redundant.
1. The search
4.6.3 Missing outcomes
The standard search of the Cochrane Schizophrenia Group Tri-
No usable outcomes were found for the following categories:
als Register (March 2007) resulted in two citations (Mahal 1976;
death, global state, general functioning, quality of life and service
Ramu 1992). We included both of these in the review. Inspection
outcomes. It is surprising that none of the included studies have
of the list of references from these two studies warranted further
reported on changes in global state.
inspection of one study (Fozdar 1962). This study was eventu-
4.6.4 Primary outcomes
ally excluded (see Characteristics of excluded studies table). An
Our primary outcome of adverse effects was reported only by
extra search conducted on AMED (Allied and Complementary
Ramu 1999 (n=78). None of our other primary outcomes were
Medicine,1985 - 2007) resulted in 23 citations. Only Ramu 1985
reported.
identified in this search was thought to be relevant and is waiting
further assessment as we have not been able to get hold of the full
text version. We felt that two studies, Farag 2003 and Weiss 1986,
Risk of bias in included studies should be mentioned in the excluded studies group (see Charac-
1. Randomisation teristics of excluded studies table). An additional web search gave
Out of the three included studies, one, Mahal 1976, was stated to links to CCRAS (Central Council for Research in Ayurveda and
be randomised. This study, however did not explicitly explain the Siddha). Correspondence with CCRAS resulted in the procure-
method of randomisation. None of the studies have categorically ment of a textbook titled “Ayurvedic Management of Unmada

(Schizophrenia)”. This book contained the entire text of the two suggestion of a difference between groups.
studies which our standard search had identified. In addition, we 2.6 Adverse effects
obtained another study (Ramu 1999) which we felt could be in- Only Ramu 1999 reported on “clinically relevant” specific ad-
cluded in this review. We also obtained part of the abstract of an- verse effects. Nausea and vomiting were found only in the brah-
other study (Roy 1964) which seems relevant and have contacted myadiyoga group as opposed to the placebo group (n=43, RR
CCRAS for the full text. We have included this last study in the 13.13 CI 0.80 to 216.30). Drowsiness was absent in either groups.
category awaiting assessment. Overall none of the adverse effects were statistically significant,
2. COMPARISON 1: AYURVEDIC HERBS versus PLACEBO although numbers are small and confidence intervals wide.
Two studies (Mahal 1976 and Ramu 1999, total n=214) com- 3. COMPARISON 2: AYURVEDIC HERBS versus ANTIPSY-
pared Ayurvedic herbs with placebo. Mahal 1976 compared two CHOTIC
types of herbs, brahmyadiyoga and tagara, with placebo. In the Mahal 1976 and Ramu 1999 compared Ayurvedic herbs with
title of the article Ramu 1999 stated that it is a comparison be- chlorpromazine (total n=120). Mahal compared two types of
tween brahmyadiyoga and tagara versus placebo but they have not herbs, brahmyadiyoga and tagara, with chlorpromazine. The title
mentioned tagara in the text of the report. of Ramu 1999 stated a comparison between brahmyadiyoga and
2.1 Leaving the studies early (any reason, adverse events, inefficacy tagara but did not mentioned tagara again anywhere in the text.
of treatment) 3.1 Leaving the studies early (any reason, adverse events, inefficacy
Overall 17% of people allocated to the brahmyadiyoga left the of treatment)
studies in the short term before the end of the trials and this Equal numbers of people left the study early from brahmyadiyoga,
compares with 22% in the placebo group (n=120, 2 RCTs, RR tagara and chlorpromazine (n=120, 2 RCTs, RR for brah-
0.77 CI 0.37 to 1.62). myadiyoga 0.91 CI 0.42 to 1.97).
2.2 Global state 3.2 Mental state (with particular reference to the positive and
None of the included studies reported usable data on changes in negative symptoms of schizophrenia)
the global state. 3.2.1 Not Improved
2.3 Mental state (with particular reference to the positive and There were no significant differences in the mental state compar-
negative symptoms of schizophrenia) isons between brahmyadiyoga and chlorpromazine in the Mahal
Mahal 1976 assessed changes in mental state using PSRS, MPQ 1976 study. Ramu 1999 however reported significantly less im-
and ayurvedic assessment tools. Ramu 1999 employed PSRS and provement in the brahmyadiyoga group (n=45, RR 1.82 CI 1.11
psychological assessment tools to observe changes in mental state. to 2.98). Mahal 1976 showed comparatively less improvement
2.3.1 Not Improved with the use of tagara when assessed using MPQ (n=68, RR 2.00
Mahal 1976 recorded usable data in changes in mental state using CI 1.25 to 3.19).
MPQ and Ayurvedic assessment and Ramu 1999 with a tool that 3.2.2 Uncooperative or did not comprehend
was not clearly described. The only statistically significant result No significant difference can be found between the usage of brah-
was found in the brahmyadiyoga group using Ayurvedic assess- myadiyoga or tagara versus chlorpromazine (n=68, 1 RCT, RR for
ment (n=68, 1 RCT, RR 0.56 CI 0.36 to 0.88, NNT 4 CI 3 to brahmyadiyoga 1.00 CI 0.27 to 3.68).
12). The overall change observed among the two groups is of no 3.2.3 Average improvement or endpoint scores - positive symp-
statistical significance. toms and negative symptoms (PSRS).
2.3.2 Uncooperative or did not respond (MPQ assessment) Data recorded on the average improvements or endpoint scores
This accounts for those participants who were not accounted for in both positive and negative symptom scores were highly skewed
during MPQ assessment. Here the trialists found no clear differ- but, if anything, tended to favour the chlorpromazine group.
ences (n=68, RR for brahmyadiyoga 2.00 CI 0.39 to 10.20). 3.3 Behaviour
2.3.3 Average improvement - positive or negative symptoms The Fergus Falls behaviour rating scale provided usable data (
The results of these comparisons could not be analysed as they Ramu 1999). The results however showed no clear differences
were highly skewed. However, findings, as reported in the trial between the effects of brahmyadiyoga and chlorpromazine.
Mahal 1976 did tend to favour brahmyadiyoga. 3.4 Psychological assessment
2.4 Behaviour The Critical flicker fusion threshold provided data that we could
The Fergus Falls behaviour rating scale recorded non-skewed data. analyse (Ramu 1999). None of the results however showed any
It showed no difference between groups (n=43, 1 RCT, WMD differences between the effects of brahmyadiyoga and chlorpro-
Fergus Falls Behaviour Rating 1.14 CI -1.63 to 3.91). mazine.
2.5 Psychological assessment 3.5 Adverse effects
Only critical flicker fusion threshold recorded data we could anal- Specific adverse effects were reported by Ramu 1999. Nausea and
yse and there were no clear differences between groups. Reaction vomiting were found only with the use of brahmyadiyoga (n=45,
time and vigilance data were much skewed but there was no clear 1 RCT, RR 20.45 CI 1.09 to 383.97, NNH 2 CI 2 to 38) while

drowsiness was present only among those who were administered Mahal 1976 cited reasons such as escape and “left against medical
chlorpromazine. The incidence of giddiness or somnolence in was advice”. There is no evidence from these studies that Ayurvedic
similar for both groups. medication is a major cause of unacceptable short term problems.
4. COMPARISON 3: AYURVEDIC TREATMENT versus AN- Mental state was measured in numerous ways. When ayurvedic
TIPSYCHOTIC assessment tools were employed, brahmyadigyoga had advantage
Ramu 1992 (n=36) compared the effects of Ayurvedic treatment, over placebo for the outcome of not improved (NNT 4 CI 3 to
in this case a complex mixture of many herbs, with those of chlor- 12). This is a finding that needs to be replicated and explained
promazine in acutely ill patients with schizophrenia. for readers of the study. We are unsure if the difference between
4.1 Leaving the studies early (any reason, adverse events, inefficacy improved and not improved has clinical meaning. It was a shame
of treatment) that no outcomes relevant to global state were reported. There
There was no difference in the number of participants leaving the was no suggestion on the Fergus Falls scale, noted for its real-
study early between those on Ayurvedic treatment and those being world pragmatism, of an effect on behaviour. There is a suggestion
administered chlorpromazine (~10%, n=36, RR 0.67 CI 0.13 to that brahmyadiyoga causes nausea and vomiting. These herbal
3.53). treatments are not innocuous, and their effects are well worth full
4.2 Mental state (with particular reference to the positive and investigation.
negative symptoms of schizophrenia)
4. COMPARISON 2: AYURVEDIC HERBS versus ANTIPSY-
The average endpoint and change scores recorded on BPRS or
CHOTIC
during Ayurvedic assessment could not be analysed as data were
skewed, but most scores do seem to favour the chlorpromazine Very few people left these short studies early, indicating perhaps
group. However, inspecting the raw scores does not suggest that that not only are Ayurvedic herbs acceptable, but so too was the
there are any great differences between the two treatment groups. trial design. Again, mental state was measured in several different
ways and for each of the herbal preparations. Results tended to
favour the short term effects of chlorpromazine over the herbs.
This does not mean that the herbs could not have a short term
DISCUSSION effect that has gone unnoticed. However, the findings from these
limited studies are not encouraging. It could be that the Ayurvedic
1. Applicability of findings
treatments do only have a positive effect in the longer term, but
All of the included studies took place in India. We do not know if we have no evidence for this. If they are being prescribed for short
the findings of this review are transferable to patients in different term benefit, evidence is lacking. Ratings on behaviour also tended
clinical settings from geographical areas distinct from the Indian to favour the chlorpromazine group. Chlorpromazine does cause
subcontinent. Even if there is an element of a placebo effect, this drowsiness so we were reassured that adverse effects were being
in itself may vary from continent to continent. well reported. Again, trials in this comparison implicated brah-
myadiyoga to cause nausea and vomiting. This is a herb that is not
All the included trials were of very short term duration. Since
without problematic effects (NNH 2 CI 2 to 38).
schizophrenia has a chronic course, there is the need for trials
conducted for longer periods of time to inform practice. People 5. COMPARISON 3: AYURVEDIC TREATMENT versus AN-
take these treatments for years. It is perfectly feasible that effects TIPSYCHOTIC
may not appear until after a considerable period. Just as conven-
One very small study (n=36) investigated the effects of Ayurvedic
tional antipsychotic drugs have a swift onset of action and little
treatment (comprising of a mixture of herbs, psycho-behavioural
evidence for their longer term effects, the converse may be true for
therapy and occupational therapy), comparing it with chlorpro-
Ayurvedic treatments.
mazine for acutely ill people with schizophrenia. Again, in the
2. Limited data context in which it was given, Ayurvedic treatment is acceptable
in the short term. All mental state measures have produced skewed
We found it disappointing that, despite considerable investment
results, but do tend to suggest that the overall Ayurvedic package
in clinical trials, no outcome data were available on death, service
held its own against chlorpromazine. This is much too small and
outcomes, general functioning, behaviour, engagement with ser-
short a trial to produce conclusive evidence upon which to base
vices, economic outcomes and cognitive functioning. The primary
care. It is hypothesis generating and it may be that for this inter-
outcome of this review was change in global state and no data were
vention more trials are justified, at least as an adjunct to modern
available on this broad, clinically meaningful, outcome.
drugs.
3. COMPARISON 1: AYURVEDIC HERBS versus PLACEBO
In the sort term, about 20% of people left both groups. Ramu
1999 did not cite any specific reasons for the attrition rates while AUTHORS’ CONCLUSIONS

Implications for practice it would have helped to clarify methodology and outcomes. None
of the included studies specified how participants were allocated
1. For people with schizophrenia
to treatment. Allocation concealment is essential for the result of
Most people given Ayurvedic medication or placebo do not show a trial to be considered valid and gives the assurance that selec-
short term mental state improvement, although there is some tion bias is kept to a minimum. Well described and tested blind-
weak suggestion that brahmyadiyoga may help more than placebo. ing could have encouraged confidence in the control of perfor-
When compared with a conventional antipsychotic such as chlor- mance and detection bias. Intention-to-treat analysis should be
promazine, Ayurvedic treatment does not seem very valuable and performed on all outcomes and all trial data should be made easily
it consistently seems to cause nausea and vomiting. There is no accessible. A minimal requirement should be that all data should,
evidence to suggest whether or not it is useful as an adjunct to at least be presented in numeric form. In addition, continuous data
antipsychotic drugs. In some cases Ayurvedic treatment may be should be presented with means, standard deviations (or standard
more acceptable, and, for very poor people, more affordable than errors) and the number of participants. Data from graphs, ’p’ val-
chlorpromazine. It is therefore a therapy option, but not one that ues of differences and statements of significant or non-significant
has been shown to be clearly effective. differences are of limited value.
2. For clinicians Failure to comply with the revised CONSORT statement, results
No convincing differences are evident when comparing Ayurvedic in both loss of data and confusion in the results, neither of which
medication with placebo or chlorpromazine and it may produce help clinicians or patients (Tharyan 2007).
unwanted gastrointestinal effects. All trials are small and short but 2. Specific
nevertheless do not really suggest any convincing clinical effect. It is
possible that real effects do exist, or that Ayurvedic treatment may We realise that data in this review are very few indeed. Studies on
be useful as an adjunct to modern drugs. In any case, concurrent any means of use of this widely prescribed acceptable treatment are
use of this ancient treatment may make taking modern drugs seem justified. However, we do feel uncomfortable with recommending
more acceptable and therefore be useful for people with psychoses. a trial of Ayurvedic herbs or treatment versus placebo. Much more
justified are trials using Ayurvedic herbs or treatments as an adjunct
3. For managers/policy makers to treatment such as chlorpromazine. We suggest a design for one
No direct data on hospital and services outcomes, satisfaction with such pragmatic, real world, randomised controlled trial in Table
care or economics are available. It is entirely possible, however, 1.
that in societies where Ayurvedic medications are more acceptable
than modern drugs, they could be used to enhance acceptability
of treatments such as chlorpromazine, which has effects on mental
state, service use and quality of life. ACKNOWLEDGEMENTS
Implications for research We thank Judith Wright for her assistance in the literature searches.
We would also like to thank Clive Adams for his invaluable guid-
1. General
ance and support. The warm and friendly environment conducive
If the recommendations of the revised CONSORT statement for the work involved in this review would not have been possible
(Moher 2001) had been anticipated by both authors and editors without the help of the Co-ordinator, Tessa Grant.
REFERENCES
References to studies included in this review and Tagara in Navonmada (acute schizophrenia). Ayurvedic
Management of Unmada (Schizophrenia). New Delhi:
Mahal 1976 {published data only} Central Council for Research in Ayurveda & Siddha, 1999:
∗
Mahal AS, Ramu NG, Chaturvedi DD. Double blind 59–76.
controlled study of brahmyadiyoga and tagara in the
management of various types of unmada (schizophrenia). Ramu 1992 {published data only}
Indian Journal of Psychiatry 1976;18(4):283–92. [: Ramu MG, Chaturvedi DD, Venkataram BS, Shankara
EMBASE 1978204625] MR, Leelavathy S, Janakiramiah N, Mukundan H,
Ramu MG, Chaturvedi DD, Venkataram BS, Shankara MR, Thomas KM, Ramachandra M, Mahal AS, Murthy NSN,
Leelavathy S, Janakiramiah N, Mukundan H, Thomas KM, Devidas KV. A controlled study on the role of classical
Ramachandra M, Mahal AS, Murthy NSN, Devidas KV. A Ayurvedic treatment in acutely ill patients with Unmada
double blind controlled study on the role of Brahmyadiyoga (schizophrenia). Ayurvedic Management of Unmada
(Schizophrenia). New Delhi: Central Council for Research API-2006
in Ayurveda & Siddha, 1999:89–100. Controller of Publications, Government of India. The
∗
Ramu MG, Venkataram BS, Mukundan H, Shankara Ayurvedic Pharmacopoeia of India. Vol. I, II, III, IV, V,
MR, Leelavathy S, Janakiramaiah N. A controlled study New Delhi: Controller of Publications, Government of
of ayurvedic treatment in the acutely ill patients with India, 2006.
schizophrenia (unmada) - rationale and results. Nimhans Bland 1997
Journal 1992;10(1):1–16. Bland JM. Statistics notes. Trials randomised in clusters.
Ramu 1999 {published data only} BMJ 1997;315:600.
Ramu MG, Chaturvedi DD, Venkataram BS, Shankara CCRAS
MR, Leelavathy S, Janakiramiah N, Mukundan H, Central Council for Research in Ayurveda & Siddha
Thomas KM, Ramachandra M, Mahal AS, Murthy NSN, [DEPTT. OF AYUSH, MINISTRY OF HEALTH
Devidas KV. A double blind controlled study on the role & FAMILY WELFARE, GOVT. OF INDIA.].
of Brahmyadiyoga and Tagara in Jirnomada ( chronic www.ccras.nic.in.
schizophrenia). Ayurvedic Management of Unmada
Divine 1992
(Schizophrenia). New Delhi: Central Council for Research
Divine GW, Brown JT, Frazier LM. The unit of analysis
in Ayurveda & Siddha, 1999:77–88.
error in studies about physicians’ patient care behaviour.
References to studies excluded from this review Journal of General Internal Medicine 1992;7(6):623–9.
Donner 2002
Farag 2003 {published data only} Donner A, Klar N. Issues in the meta-analysis of cluster
Farag NH, Mills PJ. A randomized-controlled trial of the randomized trials. Statistics in Medicine 2002;21:2971–80.
effects of a traditional herbal supplement on sleep onset
Egger 1997
insomnia. Complementary Therapies in Medicine 2003;Dec; Egger M, Smith GD, Schneider M, Minder C. Bias in
11(4):223–5. [: 0061115]
meta-analysis detected by a simple, graphical test. BMJ
Fozdar 1962 {published data only} 1997;315:629–34.
Fozdar N G, Doongaji, Bagadia V N, Vahia N S. Preliminary Gulliford 1999
Report Of An Indigenous Drug ”Acorus Calamus” In Gulliford MC, Ukoumunne OC, Chinn S. Components
Psychiatric Disorders. Indian Journal of Psychiatry 1962;4 of variance and intraclass correlations for the design of
(8):12–16. community-based surveys and intervention studies: data
Hakim 1964 {published data only} from the Health Survey for England 1994. American Journal
∗
Hakim RA. A Preliminary report on the use of Malkanguni of Epidemiology 1999;149:876–83.
with other indigenous drugs in the treatment of depression..
Higgins 2003
Indian Journal of Psychiatry 1964;6:142.
Higgins JP, Thompson SG, Deeks JJ, Altman DG.
Weiss 1986 {published data only} Measuring inconsistency in meta-analyses. BMJ 2003;327:
Weiss MG, Sharma SD, Sharma JS, Desai A, Doongaji 557–60.
DR. Traditional concepts of mental disorder among Higgins 2005
Indian psychiatric patients. Preliminary report of work in Higgins JPT, Green S (editors). Cochrane Handbook for
progress. Social Science and Medicine 1986;23(4):379–86. Systematic Reviews of Interventions 4.2.5 [updated May
[: 8509322] 2005]. The Cochrane Library. Chichester: John Wiley &
Sons, Ltd, 2005, issue Issue 3.
References to studies awaiting assessment
Hsiao 2006
Hsiao A F, Wong M D, Goldstein M S, Becerra L S, Cheng E
Ramu 1985 {published data only}
M, Wenger N S. Complementary and Alternative Medicine
Ramu MG, Venkataram BS. Manovikara (mental disorders)
Use Among Asian-American Subgroups: Prevalance,
in ayurveda. Ancient Science of Life 1985;Jan;4(3):165–73.
Predictors, and Lack of Relationship to Acculturation
[: 8507838]
and Access to Conventional Health Care. The Journal of
Roy 1964 {published data only} Alternative and Complementary Medicine 2006;12, No.10:
∗
Roy D, Mathur SM, Kaora. A double blind trial of 1003–10.
Nuvon- an Ayurvedic Drug on Chronic Schizophrenia.
Jüni 2001
Indian Journal of Psychiatry 1964;6:26.
Jüni P, Altman DG, Egger M. Systematic reviews in health
Additional references care: Assessing the quality of controlled clinical trials. BMJ
2001;323:42–6.
Altman 1996 Lad 1996
Altman DG, Bland JM. Detecting skewness from summary Lad V. An Introduction to Ayurveda. http://
information. BMJ 1996;313:1200. www.healthy.net/scr/article.asp?Id=373 1996.

Marshall 2000 Rockland 1965
Marshall M, Lockwood A, Bradley C, Adams C, Joy C, ROCKLAND LH, POLLIN W. Quantification of
Fenton M. Unpublished rating scales: a major source Psychiatric Mental Status for use with psychotic patients..
of bias in randomised controlled trials of treatments for Archives of general psychiatry 1965;12:23–28. [MEDLINE:
schizophrenia. British Journal of Psychiatry 2000;176: 14221688]
249–52. Russinova 2002
Meyer 1953 Russinova Z, Wewiorski N J, Cash D. Use of Alternative
Meyer BT, Lucero RJ. A validation study of the L-M Fergus Health Care Practices by Persons With Serious Mental
Falls Behavior Rating Scale. Journal of Clinical Psychology Illness: Perceived Benefits. American Journal of Public
1953;9(2):192–5. Health 2002;92, No.10:1600–03.
Schulz 1995
Moher 2001 Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical
Moher D, Schulz KF, Altman D. The CONSORT evidence of bias: dimensions of methodological quality
statement: revised recommendations for improving the associated with estimates of treatment effects in controlled
quality of reports of parallel-group randomized trials. JAMA trials. JAMA 1995;273:408–12.
2001;285:1987–91.
Singh 1994
Neddermeyer 2006 Singh R H. Kayachiktsa: 2 volumes. Chaukhamba
Neddermeyer D.M. Ayurvedic Medicine - Benefits. http: Surabharati Prakasana, Varanasi, 1994.
//ezinearticles.com/?Ayurvedic-Medicine---Benefits&id= Subbarayappa 2001
359445 2006. Subbarayappa B.V. The roots of ancient medicine : an
historical outline. Journal of Biosciences 2001;26:135–44.
Overall 1962
[MEDLINE: 11426049]
Overall JE, Gorham DR. The Brief Psychiatric Rating
Scale.. Psychological Reports 1962;10:799–812. Tharyan 2007
Tharyan P, Premkumar TS, Mathew V, Barnabas JP,
Ramu 1999b Manuelraj. Editorial policy and reporting of randomised
Ramu MG. Review of Previous Studies. Ayurvedic controlled trials: A systematic survey of instructions for
Management of Unmada (Schizophrenia). New Delhi: authors and assessment of trial reports in Indian medical
Central Council for Research in Ayurveda & Siddha, journals ( 2004-2005 ). The National Medical Journal of
Jawahar Lal Nehru Bhartiya Chikitsa Avum Homeopathy India [ in press ];[ in press ]:[ in press ].
Anusandhan Bhawan, 1999:49–56.
Ukoumunne 1999
Ramu 1999c Ukoumunne OC, Gulliford MC, Chinn S, Sterne JAC,
Ramu MG, Chaturvedi DD, Venkataram BS, Shankara Burney PGJ. Methods for evaluating area-wide and
MR, Leelavathy S, Janakiramiah N, Mukundan H, Thomas organistation-based intervention in health and health care:
KM, Ramachandra M, Mahal AS, Murthy NSN, Devidas a systematic review. Health Technology Assessment 1999;3(5):
KV. Appendix 2. General instructions and guidelines 1–75.
for completing the proforma. Ayurvedic Management of Walter 1999
Unmada (Schizophrenia). New Delhi: Central Council Walter G, Rey JM. The relevance of herbal treatments for
for Research in Ayurveda & Siddha, Jawahar Lal Nehru psychiatric practice. Australian and New Zealand Journal of
Bhartiya Chikitsa Avum Homeopathy Anusandhan Psychiatry 1999;33(4):482–9.
Bhawan, 1999:111–20. ∗
Indicates the major publication for the study

CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Mahal 1976
Methods Allocation: randomly allocated - no further details.

Blindness: double.
Duration: 2 months.
Design: parallel groups.
Participants Diagnosis: both schizophrenia (NIMHANS) and unmada.

N=136.
Age: mean ~27 years.
Sex: M 61, F 47 (completer data only)
Setting: hospital.
History: duration ill 2-24 months, mean duration 6 months.
Interventions 1. Tagara: dose 8gms/day month-1, 12gms/day month-2. N=27.

2. Brahmyadiyoga: dose 8gms/day month-1, 12gms/day month-2. N=27.
3. Placebo: dose 8gms/day month-1, 12gms/day month-2. N=27.
4. Chlorpromazine: dose 200mg/day month- 1, 300mg/day month-2. N=27.
All drugs given orally, four divided doses.
Outcomes Mental state: Ayurvedic assessment, MPQ, Psychotic Symptom Rating Scale.
Leaving the study early.
Unable to use -
Mental state: Spiral after effect (over 50% attrition).
Notes * Completer data only - assumed even loss between groups.
Risk of bias
Item Authors’ judgement Description
Allocation concealment? Unclear B - Unclear
Ramu 1992
Methods Allocation: unclear - 18 participants in each group.

Blindness: rater.
Duration: 28 days.
Participants Diagnosis: both schizophrenia (ICD 9) and unmada (Charaka definition).

N=36.
Age: 16-45 years, mean ~ 26.5.
Sex: not reported.
Setting: hospital.

Ramu 1992 (Continued)
History: duration ill >1month, < 10 years.
Interventions 1. Ayurvedic treatment: combined effect of shodana (cleaning) and shamana (palliative) measures &
satwawajayachikitsa (psychobehavioural therapy). N=18.
i. Snehapana: Kalyanakaghrita for 3-7 days increasing@10ml daily, daily 15ml.
ii. Ayushman 13 (Brahmyadiyoga 500mg) 2-2-2. Ayushman 14 (Nidrakarayoga) 2-0-2.
iii. Mridu abhyanga + Dhanwantarataila 2 days after snehana followed by bhaspasweda + hot water.
iv. Virechana kashayya: 75ml (with 1-2 Ichabhedi pills whenever required). Item ii stopped for 2 days, on
day of Virechana and following day.
Where symptoms did not subside by 50% at the end of 2 weeks:
i. Kalyanakaghrita in case of Vatapittonmada, Panchagavyaghrita in case of Kaphonmada - 10ml once.
ii. Item ii and Unmadagajakesarirasa 200mg + Sootasekararasa 200mg b.i.d.
iii. Item iii was given weekly 3 times.
iv. Saraswatarista 10ml and Aswagandharista 10ml b.i.d. after meals.
2. Modern treatment. N=18.
i. Chlorpromazine (Tab.300mg/day for 2 weeks and increased to 600mg/day for next 2 weeks if improve-
ment > 50%).
ii. Trihexyphenidyl HCl (Tab.2mg/day to reduce the complications of extrapyramidal symptoms).
Diazepam (Inj. or Tab. in unmanageable cases in both groups). OCT and leisure hours in both groups
Outcomes Leaving the study early.

Mental state: BPRS, Ayurvedic assessment (Manas score and Symptoms score),
Unable to use -
Psychological assessment (no usable data).
Notes Information for one person leaving chlorpromazine group - assumed even loss between groups for other
4 people
Risk of bias
Ramu 1999
Methods Allocation: unclear.

Blindness: double blind (only mentioned in title).
Duration: 75 days.
Participants Diagnosis: both schizophrenia (WHO glossary) and unmada and manah pariksha (according to Laksana
present in patients).
N=78.
Age: 20-50 years, mean ~ 32.
Sex:M 30, F 35 (no information re sex of 13 who left early).
Setting: hospital (inpatient).
History: duration ill 2-6 years.

Ramu 1999 (Continued)
Interventions 1. Brahmyadiyoga: dose 12gms/day initial 30 days, 16gms/day 31-75th day. N=23.
2. Placebo: dose 12gms/day initial 30 days, 16gms/day 31-75th day. N=20.
3. Chlorpromazine: dose 300mg/day initial 30 days, 450 mg/day 31-75th day. N=22.
All medications administered orally in 4 divided doses.
Outcomes Leaving the study early.

Mental state: Psychotic Symptom Rating Scale, psychological assessment( reaction time, critical flicker
fusion threshold and Ferguse fall’s behaviour rating scale).
Adverse effects
Notes We have assumed that the groups were evenly distributed.
Risk of bias
ICD 9 - International Classification of Diseases - 9

NIMHANS - National Institute of Mental Health And Neuro Sciences
MPQ - Multiphasic Questionnaire
BPRS - Brief Psychiatric Rating Scale
OCT- Occupational Therapy
b.i.d- twice daily
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
Farag 2003 Allocation: randomised.

Participants: people with sleep onset insomnia, not schizophrenia
Fozdar 1962 Allocation: not randomised.
Hakim 1964 Allocation: not randomised, case series.
Weiss 1986 Allocation: not randomised, case series.

DATA AND ANALYSES
Comparison 1. AYURVEDIC HERBS versus PLACEBO (all short term)
No. of No. of
Outcome or subgroup title studies participants Statistical method Effect size
1 Leaving the study early 2 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
1.1 brahmyadiyoga 2 120 Risk Ratio (M-H, Fixed, 95% CI) 0.77 [0.37, 1.62]
1.2 tagara 1 68 Risk Ratio (M-H, Fixed, 95% CI) 1.0 [0.39, 2.54]
2 Mental state: 1. Not improved 2 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
(Ayurvedic assessment)
2.2 brahmyadiyoga (MPQ 1 68 Risk Ratio (M-H, Fixed, 95% CI) 0.84 [0.60, 1.17]
assessment)
(assessment tool not clear)
2.4 tagara (Ayurvedic 1 68 Risk Ratio (M-H, Fixed, 95% CI) 0.72 [0.49, 1.05]
assessment)
2.5 tagara (MPQ assessment) 1 68 Risk Ratio (M-H, Fixed, 95% CI) 1.04 [0.79, 1.37]
3 Mental state: 2. Unco-operative 1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
or did not comprehend
4 Mental state: 3a. Average Other data No numeric data
improvement - positive
symptoms (PSRS, high=good,
skewed data)
4.1 brahmyadiyoga Other data No numeric data
4.2 tagara Other data No numeric data
5 Mental state: 3b. Average Other data No numeric data
improvement - negative
symptoms (PSRS, high=good,
skewed data)
6 Behavioiur: Average score 1 43 Mean Difference (IV, Fixed, 95% CI) 1.14 [-1.63, 3.91]
(Fergus Falls, high score = poor)
7 Psychological assessment: 1. 1 Mean Difference (IV, Random, 95% CI) Subtotals only
Critical flicker fusion threshold
(simple)
8 Psychological assessment: 2. Other data No numeric data
Reaction and vigilance (skewed
data)
8.1 Reaction time (simple) Other data No numeric data
8.2 Vigilance Other data No numeric data
9 Adverse effects 1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
9.1 drowsiness 1 43 Risk Ratio (M-H, Fixed, 95% CI) Not estimable
9.2 giddiness 1 43 Risk Ratio (M-H, Fixed, 95% CI) 2.63 [0.11, 61.05]
9.3 nausea and vomiting 1 43 Risk Ratio (M-H, Fixed, 95% CI) 13.13 [0.80, 216.30]
9.4 somnolence 1 43 Risk Ratio (M-H, Fixed, 95% CI) 2.63 [0.11, 61.05]
Comparison 2. AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term)
No. of No. of
1 Leaving the study early 2 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
2 Mental state: 1. Not improved 2 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
(Ayurvedic assessment)
2.2 brahmyadiyoga (MPQ 1 68 Risk Ratio (M-H, Fixed, 95% CI) 1.62 [0.98, 2.67]
assessment)
(assessment tool not clear)
2.4 tagara (Ayurvedic 1 68 Risk Ratio (M-H, Fixed, 95% CI) 1.38 [0.81, 2.36]
assessment)
2.5 tagara (MPQ assessment) 1 68 Risk Ratio (M-H, Fixed, 95% CI) 2.0 [1.25, 3.19]
3 Mental state: 2. Unco-operative 1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
or did not comprehend
4 Mental state: 3a.i. Average Other data No numeric data
improvement - positive
symptoms (PSRS, zero =good,
skewed data)
5 Mental state: 3a.ii. Average Other data No numeric data
endpoint score - postive
symptoms (PSRS, zero = good,
skewed data)
6 Mental state: 3b.i. Average Other data No numeric data
improvement - negative
skewed data)
7 Mental state: 3b.ii. Average Other data No numeric data
endpoint score - negative
skewed data)
8 Behavioiur: Average score 1 45 Mean Difference (IV, Fixed, 95% CI) 3.5 [-0.18, 7.18]
(Fergus Falls, high score = poor)
9 Psychological assessment: 1. 1 Mean Difference (IV, Random, 95% CI) Subtotals only
Critical flicker fusion threshold
(simple)
10 Psychological assessment: 2. Other data No numeric data
Reaction and vigilance (skewed
data)
10.1 reaction time (simple) Other data No numeric data
10.2 vigilance Other data No numeric data
11 Adverse effects 2 180 Risk Ratio (M-H, Fixed, 95% CI) 1.20 [0.48, 2.97]
11.1 drowsiness 1 45 Risk Ratio (M-H, Fixed, 95% CI) 0.11 [0.01, 1.87]
11.2 giddiness 1 45 Risk Ratio (M-H, Fixed, 95% CI) 0.48 [0.05, 4.91]
11.3 somnolence 1 45 Risk Ratio (M-H, Fixed, 95% CI) 0.96 [0.06, 14.37]
11.4 nausea and vomiting 1 45 Risk Ratio (M-H, Fixed, 95% CI) 14.38 [0.87, 237.58]
Comparison 3. AYURVEDIC TREATMENT versus ANTIPSYCHOTIC (all short term)
No. of No. of
1 Leaving the study early 1 36 Risk Ratio (M-H, Fixed, 95% CI) 0.67 [0.13, 3.53]
2 Mental state: 1a.i. Average Other data No numeric data
endpoint score - by 4 weeks
(BPRS, high=poor, skewed
data)
3 Mental state: 1a.ii. Average Other data No numeric data
endpoint score - by 4 weeks
(Ayurvedic Assessment,
high=poor, skewed data)
3.1 manas Other data No numeric data
3.2 symptoms Other data No numeric data
4 Mental state: 1b.i. Average Other data No numeric data
change score - by 4 weeks
(BPRS, high=poor, skewed
data)
5 Mental state: 1b.ii. Average Other data No numeric data
change score - by 4 weeks
(Ayurvedic Assessment,
high=poor, skewed data)
5.1 manas Other data No numeric data
5.2 symptoms Other data No numeric data

Analysis 1.1. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 1 Leaving
the study early.
Review: Ayurvedic medicine for schizophrenia
Comparison: 1 AYURVEDIC HERBS versus PLACEBO (all short term)
Outcome: 1 Leaving the study early
Study or subgroup Herbs Placebo Risk Ratio Weight Risk Ratio

n/N n/N M-H,Fixed,95% CI M-H,Fixed,95% CI
1 brahmyadiyoga
Mahal 1976 7/34 7/34 53.8 % 1.00 [ 0.39, 2.54 ]
Ramu 1999 3/26 6/26 46.2 % 0.50 [ 0.14, 1.79 ]
Subtotal (95% CI) 60 60 100.0 % 0.77 [ 0.37, 1.62 ]

Total events: 10 (Herbs), 13 (Placebo)
Heterogeneity: Chi2 = 0.74, df = 1 (P = 0.39); I2 =0.0%
Test for overall effect: Z = 0.69 (P = 0.49)
2 tagara
Mahal 1976 7/34 7/34 100.0 % 1.00 [ 0.39, 2.54 ]
Subtotal (95% CI) 34 34 100.0 % 1.00 [ 0.39, 2.54 ]

Heterogeneity: not applicable
0.1 0.2 0.5 1 2 5 10

Favours treatment Favours control

Analysis 1.2. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 2 Mental
state: 1. Not improved.
Outcome: 2 Mental state: 1. Not improved

1 brahmyadiyoga (Ayurvedic assessment)

Mahal 1976 14/34 25/34 100.0 % 0.56 [ 0.36, 0.88 ]
Subtotal (95% CI) 34 34 100.0 % 0.56 [ 0.36, 0.88 ]

2 brahmyadiyoga (MPQ assessment)
Mahal 1976 21/34 25/34 100.0 % 0.84 [ 0.60, 1.17 ]
Subtotal (95% CI) 34 34 100.0 % 0.84 [ 0.60, 1.17 ]

3 brahmyadiyoga (assessment tool not clear)
Ramu 1999 19/23 18/20 100.0 % 0.92 [ 0.72, 1.16 ]
Subtotal (95% CI) 23 20 100.0 % 0.92 [ 0.72, 1.16 ]

4 tagara (Ayurvedic assessment)
Mahal 1976 18/34 25/34 100.0 % 0.72 [ 0.49, 1.05 ]
Subtotal (95% CI) 34 34 100.0 % 0.72 [ 0.49, 1.05 ]

5 tagara (MPQ assessment)
Mahal 1976 26/34 25/34 100.0 % 1.04 [ 0.79, 1.37 ]
Subtotal (95% CI) 34 34 100.0 % 1.04 [ 0.79, 1.37 ]

0.1 0.2 0.5 1 2 5 10


state: 2. Unco-operative or did not comprehend.
Outcome: 3 Mental state: 2. Unco-operative or did not comprehend

1 brahmyadiyoga
Mahal 1976 4/34 2/34 100.0 % 2.00 [ 0.39, 10.20 ]
Subtotal (95% CI) 34 34 100.0 % 2.00 [ 0.39, 10.20 ]

2 tagara
Mahal 1976 3/34 2/34 100.0 % 1.50 [ 0.27, 8.42 ]
Subtotal (95% CI) 34 34 100.0 % 1.50 [ 0.27, 8.42 ]

0.1 0.2 0.5 1 2 5 10

Favours herbs Favours placebo
state: 3a. Average improvement - positive symptoms (PSRS, high=good, skewed data).
Mental state: 3a. Average improvement - positive symptoms (PSRS, high=good, skewed data)
Study Intervention Mean SD N Notes
brahmyadiyoga
Mahal 1976 brahmyadiyoga 7.93 9.56 27 F test undertaken.

p<0.05 in favour of brahmyadiyoga.
Mahal 1976 placebo 1.48 7.70 27
Ramu 1999 brahmyadiyoga 17.4 8.97 here high=poor
Ramu 1999 placebo 22.5 15.2
tagara

Mental state: 3a. Average improvement - positive symptoms (PSRS, high=good, skewed data) (Continued)
Mahal 1976 tagara 2.74 8.09 27 F test undertaken.

p<0.05 not significant.
Mahal 1976 placebo 1.48 7.70 27
state: 3b. Average improvement - negative symptoms (PSRS, high=good, skewed data).
Mental state: 3b. Average improvement - negative symptoms (PSRS, high=good, skewed data)
brahmyadiyoga

p<0.05 in favour of brahmyadiyoga.
Mahal 1976 placebo -0.48 8.54 27
Ramu 1999 brahmyadiyoga 17.6 13.7
tagara
Mahal 1976 tagara -2.81 7.69 27 F test undertaken.

p<0.05 not significant.
Mahal 1976 placebo -0.48 8.54 27

Analysis 1.6. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 6
Behavioiur: Average score (Fergus Falls, high score = poor).

Outcome: 6 Behavioiur: Average score (Fergus Falls, high score = poor)
Mean Mean
Study or subgroup Treatment Control Difference Weight Difference
N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI
Ramu 1999 23 26.09 (4.6) 20 24.95 (4.63) 100.0 % 1.14 [ -1.63, 3.91 ]
Total (95% CI) 23 20 100.0 % 1.14 [ -1.63, 3.91 ]

Test for subgroup differences: Not applicable
-10 -5 0 5 10
Psychological assessment: 1. Critical flicker fusion threshold (simple).

Outcome: 7 Psychological assessment: 1. Critical flicker fusion threshold (simple)
Mean Mean
Study or subgroup Brahmyadiyoga Chlorpromazine Difference Weight Difference
N Mean(SD) N Mean(SD) IV,Random,95% CI IV,Random,95% CI
Ramu 1999 23 3.38 (0.42) 20 3.08 (0.69) 0.30 [ -0.05, 0.65 ]
Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]

Test for overall effect: Z = 0.0 (P < 0.00001)
-10 -5 0 5 10

Psychological assessment: 2. Reaction and vigilance (skewed data).
Psychological assessment: 2. Reaction and vigilance (skewed data)
Study Intervention Mean S.D N Notes Order
Reaction time (simple)
Ramu 1999 Brahmyadiyoga 3.62 3.75 23
Ramu 1999 Placebo 3.94 3.49 20
Vigilance
Ramu 1999 Placebo 24.88 16.01 20
Analysis 1.9. Comparison 1 AYURVEDIC HERBS versus PLACEBO (all short term), Outcome 9 Adverse
effects.
Outcome: 9 Adverse effects
Study or subgroup Brahmyadiyoga Placebo Risk Ratio Weight Risk Ratio

1 drowsiness
Ramu 1999 0/23 0/20 Not estimable
Subtotal (95% CI) 23 20 Not estimable

Total events: 0 (Brahmyadiyoga), 0 (Placebo)
Test for overall effect: not applicable
2 giddiness
Ramu 1999 1/23 0/20 100.0 % 2.63 [ 0.11, 61.05 ]
Subtotal (95% CI) 23 20 100.0 % 2.63 [ 0.11, 61.05 ]

3 nausea and vomiting
0.1 0.2 0.5 1 2 5 10

(Continued . . . )

(. . . Continued)
Study or subgroup Brahmyadiyoga Placebo Risk Ratio Weight Risk Ratio
Ramu 1999 7/23 0/20 100.0 % 13.13 [ 0.80, 216.30 ]
Subtotal (95% CI) 23 20 100.0 % 13.13 [ 0.80, 216.30 ]

4 somnolence
Ramu 1999 1/23 0/20 100.0 % 2.63 [ 0.11, 61.05 ]
Subtotal (95% CI) 23 20 100.0 % 2.63 [ 0.11, 61.05 ]

0.1 0.2 0.5 1 2 5 10

Analysis 2.1. Comparison 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term), Outcome 1
Leaving the study early.

Comparison: 2 AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term)
Study or subgroup Herbs Chlorpromazine Risk Ratio Weight Risk Ratio

1 brahmyadiyoga
Mahal 1976 7/34 7/34 63.6 % 1.00 [ 0.39, 2.54 ]
Ramu 1999 3/26 4/26 36.4 % 0.75 [ 0.19, 3.03 ]
Subtotal (95% CI) 60 60 100.0 % 0.91 [ 0.42, 1.97 ]

Total events: 10 (Herbs), 11 (Chlorpromazine)
Heterogeneity: Chi2 = 0.11, df = 1 (P = 0.74); I2 =0.0%
2 tagara
Mahal 1976 7/34 7/34 100.0 % 1.00 [ 0.39, 2.54 ]
Subtotal (95% CI) 34 34 100.0 % 1.00 [ 0.39, 2.54 ]

0.1 0.2 0.5 1 2 5 10


Mental state: 1. Not improved.

Outcome: 2 Mental state: 1. Not improved

1 brahmyadiyoga (Ayurvedic assessment)

Mahal 1976 14/34 13/34 100.0 % 1.08 [ 0.60, 1.94 ]
Subtotal (95% CI) 34 34 100.0 % 1.08 [ 0.60, 1.94 ]

2 brahmyadiyoga (MPQ assessment)
Mahal 1976 21/34 13/34 100.0 % 1.62 [ 0.98, 2.67 ]
Subtotal (95% CI) 34 34 100.0 % 1.62 [ 0.98, 2.67 ]

3 brahmyadiyoga (assessment tool not clear)
Ramu 1999 19/23 10/22 100.0 % 1.82 [ 1.11, 2.98 ]
Subtotal (95% CI) 23 22 100.0 % 1.82 [ 1.11, 2.98 ]

4 tagara (Ayurvedic assessment)
Mahal 1976 18/34 13/34 100.0 % 1.38 [ 0.81, 2.36 ]
Subtotal (95% CI) 34 34 100.0 % 1.38 [ 0.81, 2.36 ]

5 tagara (MPQ assessment)
Mahal 1976 26/34 13/34 100.0 % 2.00 [ 1.25, 3.19 ]
Subtotal (95% CI) 34 34 100.0 % 2.00 [ 1.25, 3.19 ]

0.1 0.2 0.5 1 2 5 10


Mental state: 2. Unco-operative or did not comprehend.

Outcome: 3 Mental state: 2. Unco-operative or did not comprehend

1 brahmyadiyoga
Mahal 1976 4/34 4/34 100.0 % 1.00 [ 0.27, 3.68 ]
Subtotal (95% CI) 34 34 100.0 % 1.00 [ 0.27, 3.68 ]

2 tagara
Mahal 1976 3/34 4/34 100.0 % 0.75 [ 0.18, 3.10 ]
Subtotal (95% CI) 34 34 100.0 % 0.75 [ 0.18, 3.10 ]

0.1 0.2 0.5 1 2 5 10

Favours herbs Favours placebo
Mental state: 3a.i. Average improvement - positive symptoms (PSRS, zero =good, skewed data).
Mental state: 3a.i. Average improvement - positive symptoms (PSRS, zero =good, skewed data)
brahmyadiyoga

p<0.05 in not significant.
Mahal 1976 chlorpromazine 12.33 10.96 27

Mental state: 3a.i. Average improvement - positive symptoms (PSRS, zero =good, skewed data) (Continued)
tagara
Mahal 1976 tagara 2.74 8.09 27 F test undertaken.

p<0.05 in favour of chlorpromazine.
Mental state: 3a.ii. Average endpoint score - postive symptoms (PSRS, zero = good, skewed data).
Mental state: 3a.ii. Average endpoint score - postive symptoms (PSRS, zero = good, skewed data)
Study Intervention Mean S.D N Notes
Ramu 1999 brahmyadiyoga 17.4 8.97 here high= poor
Ramu 1999 chlorpromazine 8.14 9.56
Mental state: 3b.i. Average improvement - negative symptoms (PSRS, zero =good, skewed data).
Mental state: 3b.i. Average improvement - negative symptoms (PSRS, zero =good, skewed data)
brahmyadiyoga

p<0.05 is not significant.
tagara
Mahal 1976 tagara -2.81 7.69 27 F test undertaken.

p<0.05 in favour of chlorpromazine.

Mental state: 3b.ii. Average endpoint score - negative symptoms (PSRS, zero =good, skewed data).
Mental state: 3b.ii. Average endpoint score - negative symptoms (PSRS, zero =good, skewed data)
Study Intervention Mean S.D N Notes
Ramu 1999 chlorpromazine 10.5 8.7
Behavioiur: Average score (Fergus Falls, high score = poor).

Outcome: 8 Behavioiur: Average score (Fergus Falls, high score = poor)
Mean Mean
Study or subgroup Treatment Control Difference Weight Difference
N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI
Ramu 1999 23 26.09 (4.6) 22 22.59 (7.56) 100.0 % 3.50 [ -0.18, 7.18 ]
Total (95% CI) 23 22 100.0 % 3.50 [ -0.18, 7.18 ]

Test for subgroup differences: Not applicable
-10 -5 0 5 10

Psychological assessment: 1. Critical flicker fusion threshold (simple).

Outcome: 9 Psychological assessment: 1. Critical flicker fusion threshold (simple)
Mean Mean
Study or subgroup Brahmyadiyoga Chlorpromazine Difference Weight Difference
N Mean(SD) N Mean(SD) IV,Random,95% CI IV,Random,95% CI
Ramu 1999 23 4.07 (3.5) 22 4.05 (3.87) 0.02 [ -2.14, 2.18 ]
Subtotal (95% CI) 0 0 0.0 [ 0.0, 0.0 ]

Test for overall effect: Z = 0.0 (P < 0.00001)
-10 -5 0 5 10
Psychological assessment: 2. Reaction and vigilance (skewed data).
Psychological assessment: 2. Reaction and vigilance (skewed data)
Study Intervention Mean S.D N Notes Order
reaction time (simple)
Ramu 1999 Chlorpromazine 4.40 4.57 22
vigilance

Adverse effects.
Outcome: 11 Adverse effects
Study or subgroup Brahmyadiyoga Chlorpromazine Risk Ratio Weight Risk Ratio

1 drowsiness
Mahal 1976 0/23 4/22 56.2 % 0.11 [ 0.01, 1.87 ]
Subtotal (95% CI) 23 22 56.2 % 0.11 [ 0.01, 1.87 ]

Total events: 0 (Brahmyadiyoga), 4 (Chlorpromazine)
2 giddiness
Ramu 1999 1/23 2/22 25.0 % 0.48 [ 0.05, 4.91 ]
Subtotal (95% CI) 23 22 25.0 % 0.48 [ 0.05, 4.91 ]

3 somnolence
Ramu 1999 1/23 1/22 12.5 % 0.96 [ 0.06, 14.37 ]
Subtotal (95% CI) 23 22 12.5 % 0.96 [ 0.06, 14.37 ]

4 nausea and vomiting
Ramu 1999 7/23 0/22 6.2 % 14.38 [ 0.87, 237.58 ]
Subtotal (95% CI) 23 22 6.2 % 14.38 [ 0.87, 237.58 ]

Total (95% CI) 92 88 100.0 % 1.20 [ 0.48, 2.97 ]
Heterogeneity: Chi2 = 6.38, df = 3 (P = 0.09); I2 =53%
0.1 0.2 0.5 1 2 5 10


Analysis 3.1. Comparison 3 AYURVEDIC TREATMENT versus ANTIPSYCHOTIC (all short term),
Outcome 1 Leaving the study early.

Comparison: 3 AYURVEDIC TREATMENT versus ANTIPSYCHOTIC (all short term)
Study or subgroup Ayurvedic treatment Chlorpromazine Risk Ratio Weight Risk Ratio
Ramu 1992 2/18 3/18 100.0 % 0.67 [ 0.13, 3.53 ]
Total (95% CI) 18 18 100.0 % 0.67 [ 0.13, 3.53 ]

Total events: 2 (Ayurvedic treatment), 3 (Chlorpromazine)
0.1 0.2 0.5 1 2 5 10

Outcome 2 Mental state: 1a.i. Average endpoint score - by 4 weeks (BPRS, high=poor, skewed data).
Mental state: 1a.i. Average endpoint score - by 4 weeks (BPRS, high=poor, skewed data)
Ramu 1992 Ayurvedic treatment 11.5 7.9 18
Outcome 3 Mental state: 1a.ii. Average endpoint score - by 4 weeks (Ayurvedic Assessment, high=poor,
skewed data).
Mental state: 1a.ii. Average endpoint score - by 4 weeks (Ayurvedic Assessment, high=poor, skewed data)
manas
symptoms

Mental state: 1a.ii. Average endpoint score - by 4 weeks (Ayurvedic Assessment, high=poor, skewed data) (Continued)
Outcome 4 Mental state: 1b.i. Average change score - by 4 weeks (BPRS, high=poor, skewed data).
Mental state: 1b.i. Average change score - by 4 weeks (BPRS, high=poor, skewed data)
Ramu 1992 Ayurvedic treatment -11.9 7.2 18
Ramu 1992 Chlorpromazine -17.3 14.2 18
Outcome 5 Mental state: 1b.ii. Average change score - by 4 weeks (Ayurvedic Assessment, high=poor, skewed
data).
Mental state: 1b.ii. Average change score - by 4 weeks (Ayurvedic Assessment, high=poor, skewed data)
manas
Ramu 1992 Ayurvedic treatment -6.5 2.9 18
symptoms
Ramu 1992 Ayurvedic treatment -32 20.4 18
ADDITIONAL TABLES
Table 1. Suggested design of study
Methods Participants Interventions Outcomes Notes
Allocation: cen- Diagnosis: it may be pre- 1a. Qualtiy of life: healthy * size of study to de-
tralised sequence gener- ferred not to use diag- Ayurvedic herbs: dose days. tect a 10% difference in
ation with table of ran- nostic categories such as and choice at clinician’s Service outcomes: days improvement with 80%
dom numbers or com- DSM IV and just to in- and patient’s discretion in hospital, time attend- certainity.
puter generated code, clude those whose men- + chlorpromazine: dose ing psychiatric outpa-

Table 1. Suggested design of study (Continued)
stratified by severity of tal health problem is des- - clinician’s and patient’s tient clinic. *** Primary outcome.
illness, sequence con- ignated as schizophre- discretion. N=150 OR Satisfaction with care:
cealed till interventions nia or psychosis and un- 1b. Ayurvedic treatment patients/carers. If scales are used to mea-
assigned. In the context mada. (as a holistic package). Global state: CGI.*** sure outcome then there
of limited provision, ran- N=300. + chlorpromazine: dose Mental state: CGI, re- should be binary cut
domisation may be the Age~any. - clinician’s and patient’s lapse.** off points, defined before
only equitable way of Sex: either. discretion. N=150. Functioning: engage- study start, of clinically
distributing care. Setting: any- 2. Chlorpromazine: dose ment with services, leav- important improvement
Blindness: double - where (preferably hospi- - clinician’s and patient’s ing the study early, living
tested. tal setting ). discretion. N=150 independently.
Duration: 1 year. History: non-acute. Adverse effects: includ-
Design: parallel groups. ing mortality.
Economic out-
comes: cost-effectiveness
and cost-benefit.
WHAT’S NEW
Last assessed as up-to-date: 18 August 2007.
Date Event Description
15 February 2010 Amended Contact details updated.
HISTORY
Review first published: Issue 4, 2007
Date Event Description
11 November 2009 Amended Plain language summary updated
5 August 2009 Amended Contact details updated.
24 April 2008 Amended Converted to new review format.
19 August 2007 New citation required and conclusions have changed Substantive amendment

CONTRIBUTIONS OF AUTHORS
Vishesh Agarwal - protocol development, searching, data extraction, analyses, writing up.
Akhil Abhijnhan - protocol development, searching, data extraction, analyses, writing up.
Prakash Raviraj - protocol development, searching, data extraction, analyses, writing up.
DECLARATIONS OF INTEREST
Vishesh Agarwal - none known.
Akhil Abhijnhan - none known.
Prakash Raviraj - none known.
SOURCES OF SUPPORT
Internal sources
• Cochrane Schizophrenia Group, UK.
External sources
• No sources of support supplied
INDEX TERMS
Medical Subject Headings (MeSH)

∗ Medicine,Ayurvedic; Antipsychotic Agents [therapeutic use]; Phytotherapy [∗ methods]; Randomized Controlled Trials as Topic;
Schizophrenia [∗ drug therapy]
MeSH check words

Humans


Cochrane Ayurveda Schizophrenia

Uploaded by

Copyright:

Available Formats

Cochrane Ayurveda Schizophrenia

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cochrane Ayurveda Schizophrenia

Uploaded by

Copyright:

Available Formats

Cochrane Database of Systematic Reviews

Ayurvedic medicine for schizophrenia (Review)

Agarwal V, Abhijnhan A, Raviraj P

Agarwal V, Abhijnhan A, Raviraj P.

Ayurvedic medicine for schizophrenia (Review)

Ayurvedic medicine for schizophrenia (Review) i

Ayurvedic medicine for schizophrenia

Vishesh Agarwal1 , Akhil Abhijnhan2 , Prakash Raviraj3

Editorial group: Cochrane Schizophrenia Group.

PLAIN LANGUAGE SUMMARY

Ayurvedic medicine for schizophrenia

BACKGROUND branches: Kayachikitsa (internal medicine); Salya tantra (surgery);

Types of interventions Search methods for identification of studies

Ayurvedic medicine for schizophrenia (Review) 3

Ayurvedic medicine for schizophrenia (Review) 4

Ayurvedic medicine for schizophrenia (Review) 5

Ayurvedic medicine for schizophrenia (Review) 6

Ayurvedic medicine for schizophrenia (Review) 7

Ayurvedic medicine for schizophrenia (Review) 8

Ayurvedic medicine for schizophrenia (Review) 9

Ayurvedic medicine for schizophrenia (Review) 11

Ayurvedic medicine for schizophrenia (Review) 12

Characteristics of included studies [ordered by study ID]

Methods Allocation: randomly allocated - no further details.

Participants Diagnosis: both schizophrenia (NIMHANS) and unmada.

Interventions 1. Tagara: dose 8gms/day month-1, 12gms/day month-2. N=27.

Notes * Completer data only - assumed even loss between groups.

Item Authors’ judgement Description

Allocation concealment? Unclear B - Unclear

Methods Allocation: unclear - 18 participants in each group.

Participants Diagnosis: both schizophrenia (ICD 9) and unmada (Charaka definition).

Ayurvedic medicine for schizophrenia (Review) 13

History: duration ill >1month, < 10 years.

Outcomes Leaving the study early.

Item Authors’ judgement Description

Allocation concealment? Unclear B - Unclear

Methods Allocation: unclear.

Ayurvedic medicine for schizophrenia (Review) 14

Outcomes Leaving the study early.

Notes We have assumed that the groups were evenly distributed.

Item Authors’ judgement Description

Allocation concealment? Unclear B - Unclear

ICD 9 - International Classification of Diseases - 9

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion

Farag 2003 Allocation: randomised.

Fozdar 1962 Allocation: not randomised.

Hakim 1964 Allocation: not randomised, case series.

Weiss 1986 Allocation: not randomised, case series.

Ayurvedic medicine for schizophrenia (Review) 15

Comparison 1. AYURVEDIC HERBS versus PLACEBO (all short term)

Comparison 2. AYURVEDIC HERBS versus ANTIPSYCHOTIC (all short term)

Comparison 3. AYURVEDIC TREATMENT versus ANTIPSYCHOTIC (all short term)

Ayurvedic medicine for schizophrenia (Review) 18

Review: Ayurvedic medicine for schizophrenia

Comparison: 1 AYURVEDIC HERBS versus PLACEBO (all short term)

Outcome: 1 Leaving the study early

Study or subgroup Herbs Placebo Risk Ratio Weight Risk Ratio

Ramu 1999 3/26 6/26 46.2 % 0.50 [ 0.14, 1.79 ]