Types of studies

We include randomised controlled trials (RCTs), cohort studies and case‐control studies. We also consider observational studies to fully examine the empirical data on vaccine trials and their use in adults with cancer.

Types of participants

Adults (16 years and over) with cancer, including:

• solid malignancies treated with chemotherapy;

• haematological cancers treated or not treated with chemotherapy (since people might be immunosuppressed even without chemotherapy);

• adults with cancer, post‐autologous (up to six months after transplantation) or allogeneic (at any time) haematopoietic stem cell transplantation (HSCT).

Studies that included a mixed population of children and adults were excluded if the children population comprised more than 30% of the study cohort.

Types of interventions

We considered for inclusion studies comparing inactivated or recombinant influenza vaccines versus placebo, versus no vaccination or versus a different vaccine. We included inactivated influenza vaccine of any type, any dose and any schedule:.

• Trivalent or other

• Whole, subunit or split virion vaccine

• Adjuvanted or non‐adjuvanted

Vaccines could be matched or unmatched to circulating strains, and vaccine fit was recorded if reported. Comparisons of the same or different vaccines given during different influenza seasons or to different cancer populations were excluded. We did not include studies comparing vaccine effects in adults with cancer versus healthy adults. We excluded studies comparing vaccine dosing regimens among patients with cancer.

Types of outcome measures

Clinical outcomes were collected for a maximum follow‐up period until the end of the influenza season following vaccination. We documented the duration of follow‐up in each study. We assessed immunological response up to three months after vaccination, as defined in each study. We assessed adverse events up to two weeks after vaccination. We included studies reporting on at least one of the review‐defined outcomes (primary or secondary).

Electronic searches

For the original review we searched the Cochrane Central Register of Controlled Trials (CENTRAL, Aug 2013; Appendix 1), MEDLINE (1948 to Aug week 3 2013; Appendix 2), Embase (1980 to 2013; Appendix 3), LILACS (to Aug 2013; Appendix 4) databases. We also searched PubMed (1948 to Aug week 3 2013), combining search terms in PubMed with a highly sensitive search filter for identifying randomised controlled trials (RCTs) as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Cochrane Handbook) and with the SIGN search strategy for identifying observational studies (SIGN 2010; Appendix 5).

For this update, we searched the Cochrane Central Register of Controlled Trials (CENTRAL; 2017, Issue 4), MEDLINE (2013 to May week 2, 2017), Embase (2013 to 2017 week 21), LILACS (2013 to May 2017) databases.

We contacted the first or corresponding author of each included study for additional information, and researchers active in the field for information on unpublished studies. There were no language or publication type restrictions.

Searching other resources

We searched the following conference proceedings: Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), European Society of Clinical Microbiology and Infectious Diseases (ECCMID), Infectious Disease Conferences (IDSA), American Society of Hematology (ASH), American Society for Blood and Marrow Transplantation (ASBMT), European Group for Blood and Marrow Transplantation (EBMT), and American Society of Clinical Oncology (ASCO) between the years 2006 and 2017. We also scanned the references of all identified studies and pertinent reviews. We searched the websites of the manufacturers of influenza vaccines (May 2017). Finally, we searched for ongoing or unpublished studies in clinical trials registry databases using the https://clinicaltrials.gov/ and https://www.clinicaltrialsregister.eu/ctr‐search/search websites.

Selection of studies

We included all studies fulfilling the eligibility criteria for design, participants and interventions. We did not restrict inclusion by outcomes reported in the abstract, but obtained the full text and attempted to identify at least one of the review‐defined outcomes from the text or from author correspondence. Two review authors independently applied the inclusion criteria to all identified and retrieved articles. We documented reasons for excluding studies from the review (Characteristics of excluded studies).

Data extraction and management

Two review authors independently performed data extraction using a data extraction form. We extracted data on the following study characteristics:

• study design;

• length of the follow‐up;

• dates of study;

• location of study;

• risk of bias;,

• description of vaccines (content, timing of vaccination and antigenic match);

• description of viral circulation degree;

• description of outcomes;

• characteristics of participants: age, sex, type of malignancy, haematopoietic stem cell transplantation (HSCT), anti‐cancer treatment, expected baseline immune suppression: primarily cellular immune dysfunction, severe; primarily cellular immune dysfunction, moderate; primarily neutropenia, severe.

Assessment of risk of bias in included studies

Two review authors (RB, MP) independently assessed the risks of bias in studies fulfilling the review inclusion criteria. We contacted authors for additional information where necessary.

Measures of treatment effect

We calculated unadjusted odds ratios (ORs) with 95% confidence interval (CI) from RCTs and non‐RCTs (cohort and case‐control studies) for dichotomous data and rates (events per person‐years). For non‐RCTs, we extracted adjusted ORs with 95% CI, as available.

Unit of analysis issues

We expected that the studies might report influenza‐like illness (ILI) as episodes that occur more than once per person. In this case, we tried to extract the number of adults experiencing at least one event, however these data were not available and studies reporting on events per person‐time were analysed as rates.

Dealing with missing data

Whenever data were missing, we attempted to contact the authors of the study and request the information.

Assessment of heterogeneity

The I² statistic was calculated for each pooled estimate, in order to assess the impact of statistical heterogeneity. I² may be interpreted as the proportion of total variation among effect estimates that is due to heterogeneity rather than to sampling error, and it is intrinsically independent of the number of studies. An I² less than 30% would suggest there is little concern about statistical heterogeneity (Higgins 2002; Higgins 2003).

Assessment of reporting biases

Given the paucity of studies in each analysis we were not able to formally assess reporting biases.

Data synthesis

Due to the paucity of trials, different study designs and heterogenous reporting of events per person or per person‐time a meta‐analysis could not be performed. We present trial results in the first plots without compiling results. We had planned to stratify the analysis by the following factors, however, due to the paucity of data we could not conduct the stratified analyses:

• participant's type of malignancy and expected degree of immune dysfunction;

• degree of viral circulation;

• vaccine matching with the seasonal circulating strains.

We created 'Summary of findings' tables for the primary outcome of mortality.

We have presented the overall quality of the evidence for each outcome according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach, which takes into account issues not only related to internal validity (risk of bias, inconsistency, imprecision, publication bias) but also to external validity such as directness of results (Langendam 2013). We created Summary of findings' tables based on the methods described the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) and using GRADEpro GDT. We used the GRADE checklist and GRADE Working Group quality of evidence definitions (Meader 2014). We downgraded the evidence from 'high' quality by one level for serious (or by two for 'very serious') concerns for each limitation.

Subgroup analysis and investigation of heterogeneity

We planned to perform subgroup analyses by the expected severity of immune suppression:

• primarily cellular immune dysfunction, severe: participants post‐allogeneic haematopoietic stem cell transplantation (HSCT);

• primarily cellular immune dysfunction, moderate: chronic lymphocytic leukaemia (CLL) treated with alkylating agents, multiple myeloma (MM) treated with monoclonal antibodies;

• primarily neutropenia, severe: participants with severe neutrophil dysfunction: administration of vaccine during neutropenia (e.g. acute leukaemias, autologous HSCT, sarcoma).

Due to the paucity of data we could not perform subgroup analyses. As data accumulate in future updates we will reconsider these analyses.

Sensitivity analysis

We had planned sensitivity analyses based on studies' risk of bias for the primary outcome, but restricting the analysis to RCTs with adequate allocation generation and concealment methods, and to cohort studies at low risk of bias according to the Newcastle‐Ottawa Scale adapted for our review (NOS 2010). As previously, data did not permit such analyses.