The following study might give you some direction:

Meta-analysis of Observational Studies in Epidemiology

Before attempting to perform you have to check if the studies found are similar in relation to various items. Afterwards, a heterogeneity analysis should be performed to verify if a meta-analysis is valid.

With respect to being 1 arm study. You will be comparing it with the absence of this treatment, in this way will be evaluated if the intervention is valid. But the quality and the way the methodology of these studies have been carried out has to be kept in mind.

The Joanna Briggs Institute's web page is attached to the manuals for systematic reviews. There are still some courses that are available online.

http://joannabriggs.org/sumari.html

You can use MOOSE strategy, a derivate protocol of STROBE. 

Please let me know if the following references/sites are useful to you:

1.  Methodological Standards for Meta-Analyses and ... - Circulation

circ.ahajournals.org/content/early/2017/08/07/CIR.0000000000000523.full-text.pdf

Aug 7, 2017 - analysis and to provide recommendations that will be useful for ...

meta-analyses (fixed effect or random effects) on the basis of clinical ..... perimental, observational, or both categories of studies. ..... single-blinded studies); and statistical heterogeneity in ..... the results toward no difference between arms.[PDF]

3.  Meta-analysis methods for combining information from different ...

https://lra.le.ac.uk/bitstream/2381/28996/1/U161814.pdfSimilar

by AJ Sutton - ‎2002 - ‎Cited by 4 - ‎Related articles

... single study design (often referred to as meta-analysis), and then from multiple sources of ... publication bias, and how it can be addressed in a generalised synthesis ... observational cohort studies for both benefit and harms as well as quality of life data. ... 3.4 Combining studies with historical controls (single-arm studies).

3.  Systematic review and meta-analysis: techniques and a guide for the ...

https://www.ncbi.nlm.nih.gov › NCBI › Literature › PubMed Central (PMC)

by K Phan - ‎2015 - ‎Cited by 83 - ‎Related articles

Feb 4, 2015 - However, risk of bias can be minimized and potentially useful conclusions can ... trials available and that the evidence is mainly limited to observational studies, then .... Meta-analysis of proportions of single-arm studies reporting paravalvular ..... Article; |; PubReader; |; ePub (beta); |; PDF (550K); |; Citation ...

4.  Systematic review of methods used in meta-analyses ... - BioMedSearch

www.biomedsearch.com/attachments/00/22/55/39/22553987/1471-2288-12-64.pdfCached

by FC Warren - ‎2012 - ‎Cited by 19 - ‎Related articles

May 3, 2012 - allows the combination of individual studies, which can increase power and provide stronger evidence relating to ... observational studies only in 33.7% (56/ 166) and a mix of ..... was a meta-analysis of single-arm studies only.

Dennis

Dennis Mazur

Hi there,

I just have to say - "1-arm observational study" doesn't sound like an official thing to me. "1-arm" implies an intervention study. If you want to follow a group of people with a condition (and no healthy comparison group), I guess that would be a one-arm observational study, but most people would see that as a case-series.

If you are following a group of people who are self-intervening - such as a group of alcoholics who have decided to start going to Alcoholics Anonymous - I think it still is a case series.

But I guess if you follow them through time, it would be a cohort study technically - but you'd have to have a hypothesis - like, "Among diabetics, proteinuria precedes having a heart attack". Then you'd see if the ones with proteinuria got heart attacks at a greater rate. Or, "Among alcoholics choosing AA, exposure to AA has a dose-response association with probability of staying sober". But that's an old-fashioned observational study in a cohort with a particular condition.

I think you should probably add more information to this question. What is the condition? Do you have a few articles you are thinking of including? The literature is really different for every condition, so that really matters. People overstudy some things and understudy others.

-Monika

Dear All (Gulzar, Anibal, Wendel, Dennis and Monika)

Thank you very much for your logical and meaningful responses on my queries.

I would really appreciate your time and justification provided to explore the subject.

I was interested in pooling results of observational studies with one-arm to check difference between pooled effect estimate and simply averaged effect estimate.

For example: Polycystic ovary syndrome (PCOS) during pregnancy and the risk of fetal death

To answer my objective, I went on exploring many reviews and meta-analyses, and found below two articles that might be helpful.

Perforation in colorectal stenting: a meta-analysis and a search for risk factors reporting one-arm meta-analyses with low event rates (van Halsema et al 2014)

The author was interested in evaluating the perforation rate in colorectal stenting with self-expanding metal stents (SEMS) in patients with a colorectal obstruction. The paper concluded that the overall perforation rate of colonic stenting is 7.4% and that stent design, benign etiology of the colonic stricture, and concomitant use of chemotherapeutic agent bevacizumab were risk factors for colonic perforation.

Key statistical assumptions and methods in one-arm meta-analyses with binary endpoints and low event rates, including a real-life example in the area of endoscopic colonic stenting (Rousseau MJ 2017)

Rousseau MJ 2017 published one review article with objective of exploring the assumptions, methods, and some potential problems that can occur when performing a one-arm meta-analysis with binary data and small event rates. The authors considered a real-life motivating example taken from the endoscopic colonic stenting literature (van Halsema et al 2014), to illustrate the consequences of failure to thoroughly investigate the assumptions underlying one-arm meta-analyses, especially when there are studies that are small and have binary outcomes with zero event rates. The authors explored the impact of fixed versus random effects models and logit versus arcsine transformation on the pooled effect estimate.

The author reported below important limitations of traditional meta-analysis with log or logit transformation of binary outcome data.

• Use of an arcsine transformation instead of a logit transformation was the more appropriate method to use for a meta-analysis of literature in the area of colonic stenting as it pertains to a low incidence metric, namely colonic perforation risk.
• Use of the arcsine transformations substantially lowered the estimate for risk of colonic perforation.
• Failure to use appropriate methods may create the wrong conclusions about the risk of a treatment and could cause physicians to receive biased information on which to base selection of optimal methods of treatment.

I would suggest you to go through above two articles to explore more about one-arm meta-analysis.

Thank you again!

- Amit

You can perform single-arm analysis with R.

There is package meta with metamean for continuous outcomes and metaprop for binary outcomes.

https://cran.r-project.org/web/packages/meta/meta.pdf