Corticosteroids for preventing postherpetic neuralgia

Abstract

Background

Postherpetic neuralgia (PHN) is a common, serious, painful complication of herpes zoster. Corticosteroids have anti‐inflammatory properties, and might be beneficial. This is an update of a review first published in 2008, and previously updated in 2013.

Objectives

To assess the effects (benefits and harms) of corticosteroids in preventing postherpetic neuralgia.

Search methods

We updated the searches for randomised controlled trials (RCTs) of corticosteroids for preventing postherpetic neuralgia in the Cochrane Neuromuscular Specialised Register, CENTRAL, MEDLINE, Embase, two other databases, and two trials registers (June 2022). We also reviewed the bibliographies of identified trials, contacted authors, and approached pharmaceutical companies to identify additional published or unpublished data.

Selection criteria

We included all RCTs involving corticosteroids given by oral, intramuscular, or intravenous routes for people of all ages, with herpes zoster of all degrees of severity within seven days after onset, compared with no treatment or placebo, but not with other treatments.

Data collection and analysis

Two review authors independently identified potential articles, extracted data, assessed the risk of bias of each trial, and the certainty of the evidence. Disagreement was resolved by discussion among the co‐authors. We followed standard Cochrane methodology.

Main results

We identified five trials with a total of 787 participants that met our inclusion criteria. No new studies were identified for this update. All were randomised, double‐blind, placebo‐controlled parallel‐group studies.

The evidence is very uncertain about the effects of corticosteroids given orally during an acute herpes zoster infection in preventing postherpetic neuralgia six months after the onset of herpes (risk ratio (RR) 0.95, 95% confidence interval (CI) 0.45 to 1.99; 2 trials, 114 participants; very low‐certainty evidence (downgraded for serious risk of bias and very serious imprecision)). The three other trials that fulfilled our inclusion criteria were not included in the meta‐analysis because they did not provide separate information on the number of participants with PHN at six months.

Adverse events during or within two weeks after stopping treatment were reported in all five included trials. There were no observed differences in serious (RR 1.65, 95% CI 0.51 to 5.29; 5 trials, 755 participants; very low‐certainty evidence (downgraded for serious risk of bias and very serious imprecision)), or non‐serious adverse events (RR 1.30, 95% CI 0.90 to 1.87; 5 trials, 755 participants; low‐certainty evidence (downgraded for serious risk of bias and serious imprecision)) between the corticosteroid and placebo groups.

One of these trials was at high risk of bias because of incomplete outcome data, two were at unclear risk of bias, and the other was at low risk of bias. The review was first published in 2008; no new RCTs were identified for inclusion in subsequent updates in 2010, 2013, and 2023.

Authors' conclusions

Based on the current available evidence, we are uncertain about the effects of corticosteroids given orally during an acute herpes zoster infection on preventing postherpetic neuralgia.

Corticosteroids given orally or intramuscularly may result in little to no difference in the risk of adverse events in people with acute herpes zoster.

Some researchers have recommended using corticosteroids to relieve the zoster‐associated pain in the acute phase of the disease. If further research is designed to evaluate the efficacy of corticosteroids for herpes zoster, long‐term follow‐up should be included to observe their effect on the transition from acute pain to postherpetic neuralgia. Future trials should include measurements of function and quality of life, as well as updated measures of pain.

Plain language summary

What are the effects of corticosteroids on preventing postherpetic neuralgia?

Key messages

• We do not know if corticosteroids, given by mouth during the acute phase of a herpes zoster infection (shingles) have an effect on the prevention of postherpetic neuralgia.
• Corticosteroids given by mouth or by injection into the muscle tissue may result in little to no difference in the risk of adverse events in people with acute herpes zoster.

What is postherpetic neuralgia?

Postherpetic neuralgia (PHN) is a painful condition that can occur after a bout of shingles in the area where the rash occurred. Postherpetic neuralgia may persist until the person dies, and has major implications for quality of life and the use of healthcare resources. Many people with postherpetic neuralgia find that treatment has little or no effect at all. Therefore, attention has turned to stopping the development of postherpetic neuralgia. Some researchers propose that if a person develops PHN, it may suggest that they have more varicella‐zoster virus present in their bloodstream than the low amounts usually found when the virus is dormant; this leads to continued inflammation. Corticosteroids exert a potent anti‐inflammatory action, which might minimise nerve damage, and thereby relieve herpes zoster pain and prevent postherpetic neuralgia.

What did we want to find out?

We wanted to know if corticosteroids given acutely (within one week of the onset of the rash) during a herpes zoster infection would be effective in preventing postherpetic neuralgia. We included studies that compared corticosteroids with no treatment or placebo, but not with other treatments. We also included trials that compared corticosteroids plus routine treatment with placebo plus routine treatment.

What did we do?

We searched for all relevant randomised controlled trials that described the effects of corticosteroids on preventing postherpetic neuralgia after an acute herpes zoster infection. We also assessed how certain the evidence was, considering factors, such as study size and the way in which studies were conducted. We compared and summarised the results of the studies and rated our confidence in the evidence.

What did we find?

We included five studies examining the effects of corticosteroids on a total of 787 participants in our analysis. The evidence is very uncertain about the effects of corticosteroids, given by mouth during the acute phase of a zoster infection, in preventing postherpetic neuralgia six months after the onset of the acute herpetic rash. Corticosteroids given orally or intramuscularly may result in little to no difference in the risk of adverse events in people with acute herpes zoster. Based on the available studies, very low‐certainty evidence neither supports nor refutes the use of corticosteroids in acute herpes zoster infection for preventing postherpetic neuralgia.

What are the limitations of the evidence?

Our confidence is very limited because the results from the studies involved a limited number of people. Some studies did not clearly report the method of randomisation, allocation concealment, or whether the lack of compliance was due to inefficacy or side effects, which could have affected the study's results. Overall, we have very little confidence in the evidence, and further studies are needed to investigate the effects of corticosteroids in preventing postherpetic neuralgia.

How up‐to‐date is this review?

The evidence is up‐to‐date to 25 June 2022.

Summary of findings

Summary of findings 1. Corticosteroids for acute herpes zoster to prevent postherpetic neuralgia.

Corticosteroids for acute herpes zoster to prevent postherpetic neuralgia
Patient or population: people with acute herpes zoster (within 7 days of onset of rash) Settings: hospitals and clinics Intervention: corticosteroids Comparison: placebo or routine treatment
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Placebo or routine treatment	Corticosteroids
Presence of PHN 6 months after the onset of acute herpetic rash Clinical manifestation Follow‐up: 6 to 23 months	193 per 1000	183 per 1000 (87 to 384)	RR 0.95 (0.45 to 1.99)	114 (2 RCTs)	⊕⊝⊝⊝ Very lowa
Serious adverse events Clinical manifestation and laboratory examination Follow‐up: 6 to 23 months	8 per 1000	13 per 1000 (3 to 42)	RR 1.65 (0.51 to 5.29)	755 (5 RCTs)	⊕⊝⊝⊝ Very lowb
Non‐serious adverse events Clinical manifestation and laboratory examination Follow‐up: 6 to 23 months	113 per 1000	147 per 1000 (102 to 211)	RR 1.30 (0.90 to 1.87)	755 (5 RCTs)	⊕⊕⊝⊝ Lowc
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; PHN: postherpetic neuralgia; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded 1 level for serious risk of bias and 2 levels for very serious imprecision, due to the low number of events and participants and very wide CI.
^bDowngraded 1 level for serious risk of bias and 2 levels for very serious imprecision, due to the low number of events and very wide CI.
^cDowngraded 1 level for serious risk of bias and 1 level for serious imprecision, due to very wide CI.

Background

This is an updated version of the Cochrane review last published in 2013 (Han 2013).

Description of the condition

Postherpetic neuralgia (PHN) is a painful condition that occurs in people following an acute herpes zoster infection (commonly referred to as shingles). Shingles is an acute vesicular eruption involving one or two adjacent dermatomes (defined as an area of skin in which sensory nerves derive from a single spinal nerve root, which lies between adjacent vertebrae), with pain often preceding the eruption by days to weeks. Herpes zoster and PHN result from reactivation of the varicella‐zoster virus (VZV) acquired during chickenpox, the primary varicella infection (Gan 2013; Kost 1996; Saguil 2017; Schmader 2018; Stankus 2000; Thompson 2021). Reactivation of latent VZV from dorsal root ganglia is responsible for the classic dermatomal rash and pain that occurs with herpes zoster (Asada 2019; Harbecke 2021; Kost 1996).

Unlike varicella (chickenpox), herpes zoster is a sporadic disease with an estimated lifetime incidence of 10% to 20%. The incidence of herpes zoster increases sharply with advancing age, roughly doubling in each decade past the age of 50 years. Herpes zoster is uncommon in people younger than 15 years old (Le 2019; Stankus 2000). Previous studies have reported that people older than 55 years of age accounted for more than 30% of herpes zoster cases, while children younger than 14 years old represented only 5% of herpes zoster cases (Donahue 1995). The normal age‐related decrease in cell‐mediated immunity is thought to account for the increased incidence of VZV reactivation (Stankus 2000). People with disease states that affect cell‐mediated immunity, such as HIV infection and certain malignancies, are at increased risk. Chronic corticosteroid use, chemotherapy, and radiation therapy may increase the risk of developing herpes zoster (Fillet 2002; Kawai 2017; McKay 2020; Yamada 2023). Ethnic background may influence susceptibility to herpes zoster. Black people are one fourth less likely than white people to develop it. Although herpes zoster is not as contagious as primary varicella infection, people with reactivated infection can transmit VZV to non‐immune contacts. There is no seasonal incidence; the areas affected tend to be on the chest and abdomen, and the area controlled by the ophthalmic division of the trigeminal nerve.

PHN is one of the most common complications of herpes zoster. About 20% of people with herpes zoster develop PHN (Chawki 2022). Age is the most common risk factor, with PHN occurring nearly 15 times more frequently in people older than 50 years; the incidence of PHN after an outbreak of shingles is 10% in people over 40 years and 20% to 50% in people over 60 years (Griffin 1998; Rosler 1996; Shirane 2017; Sun 2021). Other possible risk factors for the development of PHN are ophthalmic zoster, a history of prodromal (defined as relating to the period between the appearance of initial symptoms and the full development of a rash or fever) pain before the appearance of skin lesions, and an immunocompromised state (Choo 1997; Stankus 2000). PHN may persist until death, and has major implications for quality of life and use of healthcare resources. Although PHN has been defined in different ways, previous data support the distinction between acute (within 30 days of rash onset), subacute (30 to 120 days after rash onset), and PHN (defined as pain lasting at least 120 days from rash onset (Dworkin 1994)). There is a tendency for PHN to improve with time, and as few as 3% of people are left with severe PHN after one year. However, some series report that as many as 40% of people with PHN will continue to have long‐term problems because of incomplete or no pain relief from treatments. There is no way to predict who will recover (De Moragas 1957).

Description of the intervention

Although the intensity of PHN typically decreases over time, pain lasts more than one year in more than 30% of people with shingles (Kawai 2014). Given the complexity of the underlying pathophysiological mechanisms that contribute to the development of PHN, various strategies to prevent PHN have been proposed, including corticosteroids, antidepressants and anticonvulsants, antiviral agents, and vaccination (Kim 2017). With insufficient good quality evidence, there are no recommended first‐line strategies for the treatment of PHN, except among a limited and well‐selected population.

Corticosteroids (glucocorticoids, steroids) include steroid hormones that are naturally produced in the adrenal cortex of vertebrates and their synthetic analogues. Synthetic derivatives of natural steroids include prednisone, prednisolone, methylprednisolone, betamethasone, dexamethasone, triamcinolone, and hydrocortisone. Corticosteroids are involved in many physiological processes, including stress response, immune response and regulation of inflammation, carbohydrate metabolism, protein catabolism, blood electrolyte levels, and behaviour. In addition, corticosteroids are important components in the treatment of many inflammatory, allergic, immunologic, and malignant disorders, which can be administered by inhalation, orally, or intravenously (Rhen 2005).

How the intervention might work

VZV is a highly contagious DNA virus. It is thought that the varicella virus passes to the dorsal root ganglion via the skin during the initial infection (chicken pox) and lies dormant. The treatment of PHN is quite disappointing, and efforts have turned to approaches that may prevent the development of PHN. The pathologic studies of PHN have demonstrated damage to the sensory nerves, sensory dorsal root ganglia, and dorsal horns of the spinal cord; further pathological analysis revealed diffuse and focal infiltration of chronic inflammatory cells (Mahalingam 1993; Patil 2022; Smith 1978). The presence of PHN may reflect the persistence of higher amounts of VZV than are usually found during latency, with continued inflammation. Corticosteroids exert a potent anti‐inflammatory action, which might minimise nerve damage, and thereby, relieve HZ pain and prevent PHN. On the one hand, corticosteroids effectively reduce the release of neurogenic inflammatory factors, such as substance P and calcitonin gene‐related peptide; on the other hand, they could also effectively break the vicious cycle that leads to local tissue damage and inflammation. High‐dose oral corticosteroids have been shown to produce a modest reduction in acute pain, though no effect on the incidence of PHN has been demonstrated (Whitley 1996; Wood 1994).

Some older studies designed to evaluate the effectiveness of corticosteroids, such as prednisolone, triamcinolone, or prednisone therapy in preventing PHN suggested decreased pain at three and 12 months, while other studies demonstrated no benefit (Eaglstein 1970; Keczkes 1980; Lancaster 1995; Volmink 1996). Another two large randomised, placebo‐controlled trials evaluated the combination of corticosteroids and the antiviral agent, aciclovir. One found that the addition of prednisone reduced the incidence and severity of acute pain, but provided no additional benefit for long‐term pain over aciclovir alone (Wood 1994). The other found that aciclovir and a corticosteroid did not significantly alter the course of long‐term zoster‐associated pain, but might improve quality of life (Whitley 1996). Despite the lack of clear evidence, corticosteroids are commonly used in the treatment of herpes zoster.

Why it is important to do this review

There have been a number of trials and reviews to assess the effects of corticosteroids for preventing PHN; however, their results have been inconsistent, and we still have no clear evidence. We published the first version of this systematic review of corticosteroids for preventing PHN to evaluate the evidence and synthesise available trial data. This Cochrane review is an update of the 2013 version (Han 2013). Our last review showed moderate‐certainty evidence that did not support the use of corticosteroids in acute herpes zoster infection for preventing PHN. Therefore, we undertook this update of the review to identify whether there were new randomised controlled trials that evaluated the effectiveness of corticosteroids in preventing PHN. We also revised the reporting of the review and supplemented the methods section to meet current Cochrane standards (Higgins 2023a).

Objectives

To assess the effects (benefits and harms) of corticosteroids in preventing postherpetic neuralgia.

Methods

Criteria for considering studies for this review

Types of studies

We searched for all randomised controlled trials (RCTs) of corticosteroids for preventing postherpetic neuralgia (PHN) after an acute herpes zoster infection, regardless of language or publication status. We excluded quasi‐RCTs (trials in which a partly systematic method of randomisation, such as alternation or hospital number was used).

Cluster‐randomised and cross‐over trials were not eligible for the review. Cross‐over trials were not appropriate, as there was no possibility to return to the baseline situation after the first intervention, and the medical condition evolved over time.

Types of participants

We included people of all ages with herpes zoster of all degrees of severity, within seven days after the onset.

We included trials in which participants received different treatments only if the trial also included people treated with corticosteroids. If a trial included a proportion of participants not eligible for inclusion in this review, we would have included it only if the data for eligible participants were reported separately.

Types of interventions

We searched for all kinds of corticosteroids, including hydrocortisone, prednisone, prednisolone, triamcinolone, and dexamethasone given by oral, intramuscular, or intravenous routes during the acute stage (starting within one week of the onset of the rash).

We included trials that compared corticosteroids with no treatment or placebo, but not with other treatments. We also included trials that compared corticosteroids plus routine treatment with placebo plus routine treatment.

We did not include other forms of administration of corticosteroids, such as epidural injection or topical administration, as the focus of the review was the role of systemic corticosteroids for PHN.

Types of outcome measures

We did not use the outcomes selected for the review as a basis for including or excluding studies. We excluded studies with less than one‐month follow‐up.

Primary outcomes

The primary outcome measure was the presence of PHN six months after the onset of the acute herpetic rash. PHN was defined according to clinical diagnostic criteria as pain persisting, or recurring, at the site of shingles at least one month after the onset of the acute rash (MacDonald 2000).

Secondary outcomes

1. Pain severity, measured by a validated visual analogue scale (VAS) or numerical descriptive scale, after three, six, and 12 months

2. Quality of life, measured with the Short Form‐36 Health Survey questionnaire (SF‐36), after six months (Ware 1998)

3. Adverse events during or within two weeks after stopping treatment. Adverse events were categorised as serious or not serious. Serious adverse events were those that were life‐threatening, required or prolonged hospitalisation, or caused death.

Search methods for identification of studies

We searched for all relevant randomised controlled trials (RCTs), without any language or publication restrictions.

Electronic searches

We searched the following databases.

1. Cochrane Neuromuscular Specialised Register (25 June 2022; Appendix 1)

2. Cochrane Central Register of Controlled Trials (CENTRAL; 2022, Issue 6) in the Cochrane Library (searched 25 June 2022; Appendix 2)

3. MEDLINE Ovid (January 1966 to June 2022; Appendix 3); we combined this search with the Cochrane highly sensitive search strategy (Lefebvre 2022)

4. Embase Ovid (January 1980 to June 2022; Appendix 4)

5. Chinese Biomedical Retrieval System (January 1978 to June 2022; Appendix 5)

6. LILACS (Latin American and Caribbean Health Science Information database; January 1982 to June 2022; Appendix 6)

7. ClinicalTrials.gov (searched 25 June 2022; Appendix 7)

8. World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP; searched 25 June 2022; Appendix 8)

Searching other resources

We checked the references of published studies to identify additional trials. We reviewed the bibliographies of the trials identified, contacted the study authors and known experts in the field, and approached pharmaceutical companies to identify additional published or unpublished data. We also examined any relevant retraction statements and errata.

Data collection and analysis

Selection of studies

Two review authors (X Jiang, Y Li) scrutinised the titles and abstracts identified by the searches. The review authors obtained the full text of all potentially relevant reports for independent assessment. Three review authors scrutinised all possible published and unpublished trials for inclusion. Where studies had multiple publications, we collated the reports of the same study so that each study, rather than each report, was the unit of interest for the review. In this review, we did not find multiple reports for the included studies. Publications in languages that could not be read by the authors were fully translated by a professional translator. We resolved any disagreement by discussion. We recorded the study selection process in a PRISMA flow chart (Liberati 2009).

Data extraction and management

Two review authors (X Jiang, N Chen) independently extracted the methodological details and data from publications, using a data extraction form. They extracted details and data on participants, methods, interventions, outcomes, and results. One author (X Jiang) entered the data into Review Manager (Review Manager 2020), and a second (Y Li) checked the data entry. We obtained missing data from the study authors whenever possible. We resolved disagreements by discussion.

For dichotomous outcomes, we extracted the number of events and the number of total participants in each group. For continuous data, we extracted means and standard deviations for each intervention group. When detailed data were not available, the inverse variance method was used if odds ratios and their confidence intervals were presented. For studies requiring translation, we extracted data from a translation of the study and checked numerical data against the study if possible. When numerical data were only presented in graphs, we requested the data from the study investigators. We considered using software for extracting numerical data from graphs when the data were not available elsewhere.

Assessment of risk of bias in included studies

Two review authors (X Jiang, M Zhou) independently assessed the risk of bias using the Cochrane RoB 1 tool in each trial. They classified each domain as low, unclear, or high risk of bias, as described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). The risk of bias assessments took into account security of randomisation, allocation concealment, blinding, completeness of outcome data, selective outcome reporting, and any other potential sources of bias. These items were independently assessed by two review authors. We used unclear risk of bias if there were insufficient details to assess the risk of bias, or information was available, but the risk of bias was unclear. We resolved any disagreement by discussion, with reference to a third review author (L He) if necessary. We used the overall risk of bias judgements in the GRADE approach and summary of finding tables. We considered studies at high risk of bias when any domain was assessed at high risk of bias; at unclear risk of bias if insufficient information was available with which to judge; and at low risk of bias only when all individual domains were assessed at low risk of bias.

Measures of treatment effect

We analysed the data using Review Manager software. For dichotomous outcomes, we expressed the results as risk ratios (RRs); for continuous outcomes, we compared the means and calculated mean differences (MDs), all with 95% confidence intervals (CIs). If a study used a Cox regression model for analysis, we used the Review Manager generic inverse variance feature to permit inclusion of the data in the analysis.

Unit of analysis issues

We did not include cross‐over studies or cluster‐RCTs. To avoid unit‐of‐analysis errors resulting from combining results of more than one time point for each study in a standard meta‐analysis, we evaluated outcomes based on the period of follow‐up (six months after disease onset). For studies that compared more than two intervention groups, we planned to select the relevant pair of intervention groups to include in the analyses. In studies with multiple intervention groups, we planned to combine groups if reasonable (e.g. study arms with different types of steroids). If it was not reasonable to combine the groups, we planned to split the 'shared' comparator group to avoid double‐counting of participants. No study groups were combined for the current review.

Dealing with missing data

We analysed data on an intention‐to‐treat basis, so that all participants with available data would be included in the analysis in the group to which they were allocated, regardless of whether they received the allocated intervention. If, in the original reports, participants were not analysed in the group to which they were randomised, and there was sufficient information in the trial report, we attempted to restore them to the correct group. We obtained missing data from the study authors whenever possible.

Assessment of heterogeneity

We assessed statistical heterogeneity among trials using the Chi² test with a 10% level of statistical significance (P < 0.1), and the I² statistic, along with a visual inspection of the forest plots. We followed the recommendations for the heterogeneity threshold outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2023a).

1. 0% to 40% might not be important;

2. 30% to 60% may represent moderate heterogeneity;

3. 50% to 90% may represent substantial heterogeneity;

4. 75% to 100% indicates considerable heterogeneity.

If we identified substantial heterogeneity, we reported it and explored possible causes by prespecified subgroup analysis.

Assessment of reporting biases

There were insufficient studies to use a funnel plot to investigate the possibility of publication bias. If there are sufficient studies in future updates, we will assess publication bias, as described in our protocol.

Data synthesis

We undertook meta‐analyses only when this was meaningful, that is, if the treatments, participants, and the underlying clinical question were similar enough for pooling to make sense. We originally used RevMan 5 to undertake a meta‐analysis of the studies that reported on the efficacy of corticosteroids for preventing PHN, and displayed the results as a forest plot (Review Manager 2020). Had we included a sufficient number of similar studies, we planned to compare different types of steroid versus placebo or no treatment separately. If there were insufficient studies, we included any type of steroid versus placebo or no treatment in a single comparison.

We used a fixed‐effect model for meta‐analysis unless we identified unexplained heterogeneity, in which case, we planned to use a random‐effects model analysis. For trials that were clinically heterogeneous or provided insufficient information for pooling, we undertook a descriptive analysis (Higgins 2023b).

Subgroup analysis and investigation of heterogeneity

Had there had been sufficient studies (at least 10 studies) for meaningful comparison of subgroups, we planned to use a formal statistical test in Review Manager to compare them. We planned to undertake the following subgroup analyses for the primary outcome.

1. Treatment started sooner (24 hours or less after onset) or later (24 hours to 72 hours after onset; and more than 72 hours after onset) after the onset of herpes zoster

2. Younger (up to 49 years old) and older (50 years old or more) participants

3. Type of corticosteroid used (prednisolone or triamcinolone)

Sensitivity analysis

We planned to carry out a sensitivity analysis on the basis of risk of bias, by repeating the analysis after omitting the trials that had a high or unclear risk of bias in some domains.

Summary of findings and assessment of the certainty of the evidence

We included a summary of findings table to illustrate the findings of the review and our assessment of the certainty of the evidence for the key outcomes:

1. presence of PHN six months after the onset of the acute herpetic rash;

2. serious adverse events;

3. non‐serious adverse events.

We graded the certainty of the evidence for these outcomes across the included RCTs as high, moderate, low, or very low, based on the GRADE criteria. We started at high for RCTs. Reasons for downgrading were: study limitations (risk of bias), inconsistency, imprecision, indirectness, and publication bias. We downgraded the certainty of evidence for each outcome once if a grade criterion was present to a serious degree and twice if very serious (Atkins 2004; Schünemann 2023; Skoetz 2020).

Results

Description of studies

Results of the search

We retrieved 2068 records from the comprehensive literature search, and screened 1082 records after removal of duplicates. We assessed the titles and abstracts of 1082 records, excluding 1030 records. We identified a total of 52 potentially eligible records, excluded 46 reports after full‐text screening. Five trials (reported in five references) fulfilled our inclusion criteria; we identified one record as a study awaiting classification. We identified no new studies for inclusion in this update. A flow chart illustrating study selection is shown in Figure 1.

1.

Study flow diagram

Included studies

We included five trials with a total of 787 participants in this updated review (Clemmensen 1984; Eaglstein 1970; Esmann 1987; Whitley 1996; Wood 1994a). All included trials were randomised, double‐blind, placebo‐controlled, parallel‐group studies. Two were performed in a single centre; Clemmensen 1984 was conducted in Denmark; Eaglstein 1970 was conducted in a Miami dermatology inpatient service. The others were performed in multiple centres; Esmann 1987 was conducted in Aarhus and Copenhagen, Denmark; Whitley 1996 was conducted in 15 university hospitals or affiliated clinics in the USA; and Wood 1994a was conducted in four centres in the United Kingdom.

Participants

Four trials reported a range of ages from 16 to 91 years old (Clemmensen 1984; Eaglstein 1970; Esmann 1987; Whitley 1996). Four trials stated the gender distribution (male 307, female 427), and mean age of participants (Clemmensen 1984; Esmann 1987; Whitley 1996; Wood 1994a). All five trials defined explicit inclusion criteria. One trial included participants with early, severely painful zoster (Eaglstein 1970); the other four trials included participants with herpes or pain of different grades of severity.

One trial included participants who were at least 60 years old, with onset of herpes zoster less than 96 hours before admission (Esmann 1987). One trial included immunocompetent adults older than 50 years, who fell ill less than 72 hours before study enrolment (Whitley 1996). One trial included adults who fell ill less than 72 hours before study enrolment (Wood 1994a).

All five included trials also defined explicit exclusion criteria. They excluded participants with peptic ulcer, psychosis, malignant disease, hypertension, diabetes, cardiac insufficiency, adrenocortical disease, tuberculosis, lymphomas, leukaemias, bacterial infections, pregnancy, or those who were on corticosteroid treatment. The time from onset to start of treatment was zero to seven days in four included trials (Clemmensen 1984; Esmann 1987; Whitley 1996; Wood 1994a); one trial only stated that the mean time was five days, and we were unable to obtain more details.

Interventions

The treatment regimens varied between studies (see Characteristics of included studies).

Two trials compared corticosteroids with placebo (Clemmensen 1984; Eaglstein 1970). Three trials used aciclovir in combination with corticosteroids versus aciclovir in combination with placebo (Esmann 1987; Whitley 1996; Wood 1994a). One trial used oral triamcinolone (Eaglstein 1970), four trials used oral prednisone (Clemmensen 1984; Esmann 1987; Whitley 1996; Wood 1994a), and one trial administered adrenocorticotropic hormone (ACTH) intramuscularly (Clemmensen 1984). All trials followed the participants for at least six months or until the PHN ended, or the participant no longer returned. One trial monitored all participants for three years (Eaglstein 1970).

Outcomes measured

Four trials reported the duration of PHN or presence of PHN at six months after the onset of the acute herpetic rash (Eaglstein 1970; Esmann 1987; Whitley 1996; Wood 1994a). Two trials did not provide separate information on the number of participants with PHN at six months and therefore, were not included in the meta‐analysis (Whitley 1996; Wood 1994a). All five included trials did not report, as a separate outcome, pain severity measured by a validated visual analogue scale or numerical descriptive scale after 3, 6, and 12 months, or the quality of life measured with the SF‐36 after six months. All five included trials reported adverse events during treatment or within two weeks after stopping treatment. Adverse events were categorised as serious or not serious.

Excluded studies

We excluded 46 trials (see Characteristics of excluded studies). The main reasons for exclusion were as follows:

1. Study design: after contacting the study authors, 20 studies were not RCTs (Agarwal 1991; Brusco 1993; Cui 2002; Hao 2002; Huang 2004; Huang 2010; Jiang 2008; Li 2000; Li 2002; Lin 2005; Ma 2000; Ma 2002; Shi 2008; Song 2009; Tang 2004; Yang 2010; Zeng 2011; Zhang 2003; Zhou 2000; Zhou 2008);

2. Concomitant treatments: in seven trials, participants received therapy that was mismatched in dosage, course of treatment, or routine treatment between corticosteroids and control groups (Guo 2001; Jiang 2005; Yang 2000; Yin 2004; Yin 2005; Zhang 2004; Zheng 2004);

3. Definition of PHN: four trials defined PHN as pain persisting at the site of shingles two weeks after the onset of the acute rash (Chang 2004; Liu 2003; Liu 2005; Yang 2002), one trial defined PHN as pain persisting one week after total decrustation (Liao 2005);

4. Short‐term follow‐up: five trials followed up for less than one month (Chang 2004; Liao 2005; Liu 2003; Liu 2005; Yang 2002);

5. Timing: four trials exceeded our seven days in treatment criterion (Lin 2002; Wang 2004; Xu 1999; Zhang 2005);

6. Concomitant anaesthetics: in six trials, the intervention group was treated with corticosteroids plus anaesthetics (Cui 2017; Ji 2009; Makharita 2012; Ni 2017; Pasqualucci 2000; van Wijck 2006);

7. Administration route: other forms of administration of corticosteroids, such as epidural injection or intrathecal injection were used in five trials (Dureja 2010; Pasqualucci 2000; Rijsdijk 2013; van Wijck 2006; NCT05208918).

Studies awaiting classification

The study currently awaiting classification did not clarify the exact course of disease from onset of herpes zoster to receipt of treatment (Hu 2001). We contacted the study authors by mail but failed to receive any reply. Another article, which was classified as a study awaiting assessment in the first version of this Cochrane review, was a review of case reports and previous trials (Levinson 1985). Although the authors of Levinson 1985 said they were investigating the feasibility of a multicentre trial in this field, we have not found any subsequent relevant trial reports, so we removed this article from studies awaiting assessment and included it under additional references.

Ongoing studies

We found no ongoing studies that matched our eligibility criteria.

Risk of bias in included studies

For this update, according to the RoB 1 assessment (Higgins 2017), we rated two trials at low risk of bias (Eaglstein 1970; Wood 1994a), two at unclear risk of bias (Clemmensen 1984; Whitley 1996), and one at high risk of bias (Esmann 1987). See Characteristics of included studies and Figure 2.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study

Allocation

Random sequence generation

Three included trials reported the method of randomisation, and we judged them at low risk of bias for random sequence generation (Eaglstein 1970; Whitley 1996; Wood 1994a). Eaglstein 1970 used randomisation with a centralised code generated at a pharmacy. A supply of medication for each participant was assigned a different code number and distributed to the participants; the code of each participant was opened after a three‐year follow‐up. Both Whitley 1996 and Wood 1994a used a computer‐generated randomisation code to randomly assign participants.

Allocation concealment

We judged Wood 1994a and Eaglstein 1970 at low risk of bias for allocation concealment. In Wood 1994a, the randomisation code was stratified by study centre to assign participants in blocks of eight to either group, so that allocation concealment might be performed. Eaglstein 1970 stated that "the code for each participant was opened after follow‐up for evaluation". Therefore, we rated this study at low risk of bias for allocation concealment. The method of allocation concealment in Whitley 1996 was unclear; therefore, we rated this domain at unclear risk of bias.

Clemmensen 1984 and Esmann 1987 did not describe the method of randomisation, and it was not clear from the reports if there was adequate allocation concealment in either study; we judged them at unclear risk of bias for both selection bias domains.

Blinding

Blinding of participants and personnel

We assessed the five included trials at low risk of performance bias. A placebo‐controlled design was used in all five trials to ensure that participants were unaware of their group allocation during the study. One of these trials used lactose as placebo (Eaglstein 1970), another used calcium lactate (Esmann 1987); the other three studies only stated that the placebo tablets were indistinguishable from the active medication, but did not describe their composition.

Blinding of outcome assessors

Eaglstein 1970, reported to be double‐blind, did not describe whether outcome assessors were blinded but stated that the code of each participant was opened after follow‐up and evaluation, so we judged it at low risk of detection bias. The other four studies clearly stated that a double‐blind design was used, and outcome assessors were blinded to treatment assignment during the entire study period. We rated these studies at low risk of detection bias.

Incomplete outcome data

In Esmann 1987, six participants were withdrawn, but it was unclear whether the lack of compliance was due to inefficacy or side effects. The group to which those participants were first assigned was not specified, so bias from incomplete outcome data was possibly an issue in this study, and we judged it at high risk of bias. The other four studies all reported information about follow‐up and dropouts and described the reasons for dropping out clearly; we considered them at low risk. Only one study claimed that an intention‐to‐treat analysis was used (Whitley 1996); the other four included studies did not state whether the analysis was by intention‐to‐treat, but there was sufficient information in three trial reports to restore them to the correct group and perform an intention‐to‐treat analysis in our review (Clemmensen 1984; Eaglstein 1970; Wood 1994a).

Selective reporting

Since four of the included trials were designed primarily to assess the efficacy of corticosteroids during the acute phase of herpes zoster infection, and all reported negative results or marginal positive results for corticosteroids for preventing PHN, we considered the reporting bias introduced by the investigators to be small (Clemmensen 1984; Eaglstein 1970; Whitley 1996; Wood 1994a). For each included study, outcomes listed in the methods section were all reported. All the study's prespecified outcomes were reported as per protocol, so we judged all five included trials at low risk for selection bias.

Other potential sources of bias

All trials considered baseline clinical features. In four trials, the baseline clinical features were similar between groups (Clemmensen 1984; Esmann 1987; Whitley 1996; Wood 1994a). Eaglstein 1970 did not describe baseline differences, but the authors reported no serious imbalances in baseline prognostic factors between groups.

The included studies used different cutoff times to definite PHN. Although we clearly stated that PHN was pain persisting, or recurring, at the site of shingles at least one month after the onset of the acute rash, we did not restrict inclusion to studies using the same definition, so we would not introduce more missing data.

We deemed all the included studies to be at low risk of other bias.

Effects of interventions

See: Table 1

See Table 1.

Corticosteroids versus placebo or no treatment

Primary outcome

Presence of PHN six months after the onset of the acute herpetic rash

One trial, comparing triamcinolone with placebo, provided data on the presence of PHN six months after the onset of the acute herpetic rash (Eaglstein 1970). Another trial comparing prednisolone plus routine treatment with placebo plus routine treatment provided data for our primary outcome (Esmann 1987). The trial compared aciclovir plus corticosteroids with aciclovir plus placebo. We conducted a meta‐analysis combining relevant data from the above two trials, with a total of 114 participants, and we are uncertain about the effects of corticosteroids given acutely during zoster infection by oral administration in preventing postherpetic neuralgia (risk ratio (RR) 0.95, 95% confidence interval (CI) 0.45 to 1.99, P = 0.89; 2 trials, 114 participants; very low‐certainty evidence; Analysis 1.1). We downgraded the certainty of the evidence by one level for serious risk of bias and two levels for very serious imprecision, due to a low number of events and participants and very wide CIs (Table 1).

1.1. Analysis.

Comparison 1: Corticosteroids versus placebo, Outcome 1: The presence of PHN six months after the onset of acute herpetic rash

Clemmensen 1984 used a cutoff time of six weeks for defining PHN, concluding that prednisone did not decrease the incidence of PHN. Since the numbers of participants with PHN six months after the rash onset could not be obtained, these data were not included in any meta‐analysis for effects of corticosteroids.

Two other trials provided relevant data for this outcome, although not in a format which permitted inclusion in our meta‐analysis. In a trial with 201 participants, a Cox regression model analysis of the main effect of prednisone compared with no prednisone showed little or no difference in the time to cessation of zoster‐associated pain (ZAP; RR 1.26, 95% CI 0.91 to 1.75 (Whitley 1996)). From their Cox regression model, we used the generic inverse variance approach to calculate the main effect of prednisone compared with no prednisone and confirmed that there was no observed difference in the time to cessation of ZAP (RR 1.11, 95% CI 0.96 to 1.27; 1 trial, 201 participants; very low‐certainty evidence; Analysis 1.2). We downgraded the certainty of evidence by one level for unclear risk of bias and two levels for very serious imprecision due to few events and very wide CIs. In the trial with 400 participants, 349 of whom completed the study, the investigators did not detect differences between any of the treatment groups in the time to complete cessation of pain (Wood 1994a). The median time to cessation of pain was 147 and 120 days in the groups receiving 7‐day and 21‐day aciclovir without a corticosteroid, and 146 and 120 days in the two groups who received aciclovir with corticosteroids. Thus, the results of both these large trials agreed with the conclusion from the meta‐analysis of the two smaller trials that corticosteroids may not affect the presence of PHN after six months.

1.2. Analysis.

Comparison 1: Corticosteroids versus placebo, Outcome 2: The main effect of prednisone compared with no prednisone on six months evaluation of pain (generic inverse variance)

Secondary outcome measures

Pain severity

Four trials evaluated pain intensity after using corticosteroids (Clemmensen 1984; Eaglstein 1970; Esmann 1987; Wood 1994a).

Three of them measured pain during the first month, using different evaluation methods, results of which we were unable to pool, even after attempting to convert the outcomes to dichotomous data. Clemmensen 1984 (40 participants) graded pain from zero (no pain) to three (insufferable pain). There was no difference in mean pain score between the prednisone and placebo groups during the 21‐day treatment period; the score was lower in the ACTH group during the first four days of the trial (P = 0.02 to 0.03), but not after. Eaglstein 1970 (35 participants) reported that in participants older than 60 years, pain resolved spontaneously, but more rapidly with corticosteroids. Wood 1994a (400 participants) observed a larger reduction in pain score in the corticosteroids group than in the non‐corticosteroids group on days 7 and 14 (P < 0.01).

Only one trial (18 participants) reported pain at six months: 18 participants reported pain, 15 of whom reported light pain, two reported moderate pain, and one reported severe pain. Of these, nine participants received prednisone and nine received placebo, but pain severity was not reported by intervention (Esmann 1987).

None of the studies reported pain at three or 12 months.

Quality of life

None of the trials reported separate data on quality of life after six months, measured with the SF‐36.

Adverse events during or within two weeks after stopping treatment

Adverse events were categorised as serious or not serious. Serious adverse events were those which were life‐threatening, required or prolonged hospitalisation, or caused death. Details for individual studies are in the Characteristics of included studies.

Serious adverse events

Two of the included trials explicitly recorded the absence of serious adverse effects attributable to the experimental treatment (Clemmensen 1984; Eaglstein 1970).

The other three trials reported several serious adverse events, including acute cardiac insufficiency, myocardial infarction, pneumonia or bronchopneumonia, chest infection, haematemesis, and death from other, unspecified causes.

The evidence is very uncertain about the effect of corticosteroids on serious adverse events (RR 1.65, 95% CI 0.51 to 5.29, P = 0.40; 5 trials, 755 participants; very low‐certainty evidence; Analysis 1.3). We downgraded the certainty of evidence by one level for serious risk of bias, and two levels for very serious imprecision due to a low number of events and very wide CIs.

1.3. Analysis.

Comparison 1: Corticosteroids versus placebo, Outcome 3: Serious adverse events

Non‐serious adverse events

All five trials reported details of non‐serious adverse events, including clinical manifestation or laboratory results, the number of participants experiencing each adverse event, and their distribution between groups. The most frequently reported non‐serious adverse events were gastrointestinal symptoms (such as dyspepsia, nausea, vomiting, and diarrhoea), dizziness, headache, sweats, rash, oedema, hyperglycaemia, and increased serum aspartate glutamyltransferase. Corticosteroids may have little or no effect on the overall incidence of non‐serious adverse events compared to placebo (RR 1.30, 95% CI 0.90 to 1.87, P = 0.16; 5 trials, 755 participants; low‐certainty evidence; Analysis 1.4). We downgraded the certainty of evidence by one level for serious risk of bias, and one level for serious imprecision due to very wide CIs.

1.4. Analysis.

Comparison 1: Corticosteroids versus placebo, Outcome 4: Non‐serious adverse events

Subgroup analyses

We did not conduct any subgroup analyses due to the paucity of studies included in each analysis, and unavailable data.

Sensitivity analyses

No significant heterogeneity (I² > 50%) was present in any analysis. We were unable to undertake sensitivity analyses since most studies were at high or unclear risk in at least one domain.

Discussion

Summary of main results

Only five studies examining the prevention effects of corticosteroids in a total of 787 participants were suitable for inclusion in this review. This is a relatively small number in relation to the known variability in the outcome of postherpetic neuralgia (PHN). Based on a limited number of studies, we are uncertain about the effects of corticosteroids, given orally, during an acute zoster infection, on preventing postherpetic neuralgia. There was no difference in time to cessation of pain, measured in two large trials (550 participants). Four of the trials included in this review evaluated pain intensity changes after using corticosteroids, but used different pain evaluation methods, so we were unable to combine the data in our meta‐analysis. We did not identify any data on quality of life measured with the SF‐36 Health Survey after six months. The evidence regarding the occurrence of serious adverse events is very uncertain. Corticosteroids may result in little to no difference in the risk of non‐serious adverse events in people with acute herpes zoster.

There was a lack of sufficient detail to permit extraction of all required data concerning subgroups of interest. In the absence of an observed effect in the primary outcome measures, and the fact that individual trials were too small to detect moderate effects, more extensive subgroup analysis would have been unreliable. We hope that publication of this review will encourage authors of future PHN prevention trials to collect and publish data that will allow the analysis of subgroups in subsequent systematic reviews.

Overall completeness and applicability of evidence

Most of the included trials addressed relevant endpoints for PHN and adverse effects, but data on the severity of pain measured by validated visual analogue or numerical descriptive scales and quality of life measured with the SF‐36 were limited. Therefore, there is no evidence to address the efficacy for these outcomes.

No other forms of administration, such as epidural injection or intrathecal administration were included. Thus, the results of our review are limited to oral or intramuscular corticosteroids.

The studies were mostly from Europe and the USA, and all the studies were over 30 years old; best practice in PHN prevention has changed over time. Some newer studies have reported that combined corticosteroids and anaesthetic can reduce the risk of PHN and duration of neuralgic pain in people with herpes zoster rash (Kim 2017; Kim 2021; Zhang 2022).

Quality of the evidence

All included trials were randomised, double‐blind, placebo‐controlled parallel studies. We rated two trials at low risk of bias, because of their design and efforts to avoid selection, performance, detection, attrition, reporting, and other biases. We rated two trials as unclear, since they had an unclear risk of bias for one or more key domains, commonly for allocation concealment and random sequence generation. We rated one trial at high risk of bias, mainly due to its incomplete outcome data.

We rated the certainty of evidence for the primary outcome as very low, due to serious risk of bias and very serious imprecision. Hence, we were unable to draw conclusions on the efficacy of corticosteroids for preventing PHN.

Potential biases in the review process

We attempted to comprehensively search the pertinent literature, including published and unpublished studies, without any language restrictions, and contacted the investigators to acquire additional information, as required. However, we were unable to obtain further data from the trial authors or pharmaceutical companies, and it remains possible that important studies were missed. To reduce the risk of attrition bias, we used an intention‐to‐treat approach for the meta‐analyses. In addition, the small number of studies did not permit any of our planned subgroup and sensitivity analyses, and questions around participant selection and optimal treatment schedule remain.

Agreements and disagreements with other studies or reviews

Because we assessed the evidence to be of low and very low certainty, we are uncertain whether corticosteroids given during an acute zoster infection have a clinical impact on the prevention of postherpetic neuralgia.

There have been a number of reviews and meta‐analyses completed for the treatment or prevention of PHN, largely based on the same trials that we included in our review, but their results were inconsistent. Naldi 1990 published an overview of four clinical trials with corticosteroids versus placebo or carbamazepine for preventing PHN. They found that corticosteroids had no effect on the development of PHN. However, a meta‐analysis of the same four trials undertaken in Lycka 1990 showed a significant decrease in the prevalence of PHN at six and 12 months amongs corticosteroids recipients. Recently, Kim 2021 undertook a comprehensive review and meta‐analysis to investigate the effects of various strategies used to prevent PHN; the results showed no benefit of corticosteroids for reducing the incidence of PHN.

Authors' conclusions

Implications for practice.

Based on the current available evidence, we are uncertain about the effects of corticosteroids given orally, during an acute herpes zoster infection, on preventing postherpetic neuralgia. Corticosteroids given orally or intramuscularly may result in little to no difference in the risk of adverse events in people with acute herpes zoster.

Implications for research.

Large trials with sufficient power to detect a meaningful difference, and which include validated and approved pain outcomes have not been performed. High‐quality studies to assess the effects of corticosteroids on both short‐term pain and longer‐term postherpetic neuralgia are required. This may provide information about the mechanisms of transition from acute pain to long‐term postherpetic neuralgia. Randomised studies are also needed to definitively affirm or rule out a clinically useful role of corticosteroids in treating acute pain and pain associated with postherpetic neuralgia.

What's new

Date	Event	Description
5 December 2023	New search has been performed	Search updated to 25 June 2022
5 December 2023	New citation required and conclusions have changed	Updated searches integrated and conclusions changed from previous update. Updated methods and reporting according to current Cochrane standards. We also updated the background and regraded the overall certainty of the evidence using GRADE. No new studies were included in this update and the evidence was very low‐ to low‐certainty. We are uncertain about the effects of corticosteroids given acutely during zoster infection in preventing postherpetic neuralgia. We made changes to the team of authors.

History

Protocol first published: Issue 1, 2006
Review first published: Issue 1, 2008

Date	Event	Description
23 October 2012	New citation required but conclusions have not changed	Updated searches integrated but no new trials identified. Change in first author.
8 November 2010	New citation required but conclusions have not changed	New authors joined the review update team
17 July 2010	New search has been performed	For the 2010 update, we updated the searches, but found no new trials. We assessed risk of bias using the new methods, added a summary of findings table and revised the review.
14 November 2007	New citation required and conclusions have changed	Substantive amendment

Appendices

Appendix 1. Cochrane Neuromuscular Specialised Register search strategy

#1 MeSH DESCRIPTOR Herpes Zoster AND INREGISTER
#2 "herpes zoster" AND INREGISTER
#3 MeSH DESCRIPTOR Neuralgia, Postherpetic AND INREGISTER
#4 "postherpetic neuralgia" AND INREGISTER
#5 #1 or #2 or #3 or ＃4 AND INREGISTER
#6 "adrenal cortex hormones" AND INREGISTER
#7 "glucocorticoid" AND INREGISTER
#8 "corticosteroid" AND INREGISTER
#9 "steroid" AND INREGISTER
#10 "prednisolone" AND INREGISTER
#11 "triamcinalone" AND INREGISTER
#12 "dexamethasone" AND INREGISTER
#13 "triamcinolone" AND INREGISTER
#14 "hydrocortisone" AND INREGISTER
#15 "prednisone" AND INREGISTER
#16 #6 or #7 or #8 or ＃9 or #10 or #11 or #12 or ＃13 or #14 or #15 AND INREGISTER
＃17 #5 AND #16 AND INREGISTER

Appendix 2. CENTRAL search strategy

#1 MeSH descriptor Herpes Zoster explode all trees
#2 “herpes zoster”
#3 shingle*
#4 MeSH descriptor Neuralgia, Postherpetic, this term only
#5 PHN
#6 (postherpetic or post herpetic or post‐herpetic) and neuralgia
#7 (#1 OR #2 OR #3 OR #4 OR #5 OR #6)
#8 MeSH descriptor Glucocorticoids explode all trees
#9 glucocorticoid*
#10 MeSH descriptor Adrenal Cortex Hormones explode all trees
#11 adrenal cortex hormone*
#12 corticosteroid*
#13 MeSH descriptor Steroids explode all trees
#14 steroid*
#15 MeSH descriptor Prednisolone explode all trees
#16 prednisolone*
#17 MeSH descriptor Triamcinolone explode all trees
#18 triamcinalone*
#19 MeSH descriptor Dexamethasone explode all trees
#20 dexamethasone*
#21 triamcinolone*
#22 MeSH descriptor Hydrocortisone explode all trees
#23 hydrocortisone*
#24 MeSH descriptor Prednisone explode all trees
#25 prednisone*
#26 (#8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20
OR #21 OR #22 OR #23 OR #24 OR #25)
#27 #7 and #26

Appendix 3. MEDLINE Ovid search strategy

1. exp Herpes Zoster/
2. herpes zoster.mp.
3. shingle$.mp.
4. neuralgia/
5. (postherpetic or post herpetic).mp.
6. 4 and 5
7. PHN.tw.
8. postherpetic neuralgia.mp.
9. post herpetic neuralgia.mp.
10. post‐herpetic neuralgia.mp.
11. or/1‐3,6‐10
12. Glucocorticoids/
13. glucocorticoid$.mp.
14. adrenal cortex hormone/
15. adrenal cortex hormone$.mp.
16. corticosteroid$.mp.
17. exp Steroids/
18. steroid$.mp.
19. Prednisolone/
20. prednisolone$.mp.
21. TRIAMCINOLONE/
22. triamcinalone$.mp.
23. DEXAMETHASONE/
24. dexamethasone$.mp.
25. triamcinolone$.mp.
26. HYDROCORTISONE/
27. hydrocortisone$.mp.
28. PREDNISONE/
29. prednisone$.mp.
30. or/12‐29
31. randomised controlled trial.pt.
32. controlled clinical trial.pt.
33. randomised controlled trials/
34. random allocation/
35. double‐blind method/
36. single‐blind method/
37. or/31‐36
38. animals/ not humans/
39. 37 not 38
40. clinical trial.pt.
41. exp clinical trial/
42. (clin$ adj25 trial$).ti,ab.
43. ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or mask$)).ti,ab.
44. placebos/
45. placebo$.ti,ab.
46. random$.ti,ab.
47. research design/
48. or/40‐47
49. 48 not 38
50. 49 not 39
51. comparative study/
52. exp evaluation studies/
53. follow up studies/
54. prospective studies/
55. (control$ or prospectiv$ or volunteer$).ti,ab.
56. or/51‐55
57. 56 not 38
58. 57 not (39 or 50)
59. 39 or 50 or 58
60. 11 and 30 and 59

Appendix 4. Embase Ovid search strategy

1. Randomized Controlled Trial/
2. Clinical Trial/
3. Multicenter Study/
4. Controlled Study/
5. Crossover Procedure/
6. Double Blind Procedure/
7. Single Blind Procedure/
8. exp RANDOMIZATION/
9. Major Clinical Study/
10. PLACEBO/
11. Meta Analysis/
12. phase 2 clinical trial/ or phase 3 clinical trial/ or phase 4 clinical trial/
13. (clin$ adj25 trial$).tw.
14. ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or mask$)).tw.
15. placebo$.tw.
16. random$.tw.
17. control$.tw.
18. (meta‐analys$ or systematic review$).tw.
19. (cross‐over or factorial or sham or dummy).tw.
20. ABAB design$.tw.
21. or/1‐20
22. human/
23. nonhuman/
24. 22 or 23
25. 21 not 24
26. 21 and 22
27. 25 or 26
28. exp *Herpes Zoster/
29. herpes zoster.tw.
30. shingle$.tw.
31. neuralgia/
32. (postherpetic or post herpetic).tw.
33. 31 and 32
34. PHN.tw.
35. postherpetic neuralgia.tw.
36. post herpetic neuralgia.tw.
37. post‐herpetic neuralgia.tw.
38. or/28‐30,33‐37
39. Glucocorticoid/
40. glucocorticoid$.tw.
41. adrenal cortex hormone/
42. adrenal cortex hormone$.tw.
43. corticosteroid$.tw.
44. exp Steroid/
45. steroid$.mp.
46. Prednisolone/
47. prednisolone$.tw.
48. TRIAMCINOLONE/
49. triamcinalone$.tw.
50. DEXAMETHASONE/
51. dexamethasone$.tw.
52. triamcinolone$.tw.
53. HYDROCORTISONE/
54. hydrocortisone$.tw.
55. PREDNISONE/
56. prednisone$.tw.
57. or/39‐56
58. 27 and 38 and 57

Appendix 5. Chinese Biomedical Retrieval System Database search strategy

(NB. all of the search terms were translated to Chinese terms when we conducted the searches)
1. herpes zoster
2. postherpetic neuralgia
3. PHN
4. shingle
5. 1‐4/or
6. herpes
7. neuralgia
8. 6 and 7
9. 5 or 8
10.corticorsteroid
11.hormone
12.steroid
13.prednisone
14.prednisolone
15.triamcinolone
16.dexamethasone
17.hydrocortisone
18.10‐17/or
19.random
20.control
21.clinical trial
22.blind procedure
23.placebo
24.19‐23/or
25.9 and 18 and 24

Appendix 6. LILACS search strategy

(Mh Herpes Zoster OR Tw herpes zoster OR Tw shingle$ OR (Mh neuralgia AND (postherpetic OR post herpetic)) OR Tw PHN OR Tw postherpetic neuralgia OR Tw post herpetic neuralgia OR post‐herpetic neuralgia) AND (Mh Glucocorticoids OR Tw glucocorticoid$ ORMh adrenal cortex hormone OR Tw adrenal cortex hormone$ OR Tw corticosteroid$ ORMh Steroids OR Tw steroid$ OR Mh Prednisolone OR Tw prednisolone$ OR Mh TRIAMCINOLONE OR Tw triamcinalone$ OR Mh DEXAMETHASONE Or Tw dexamethasone$ OR Tw triamcinolone$ OR Mh HYDROCORTISONE OR Tw hydrocortisone$ OR Mh PREDNISONE OR Tw prednisone$) AND ((Pt randomised controlled trial OR Pt controlled clinical trial OR Mh randomised controlled trials OR Mh random allocation OR Mh double‐blind method OR Mh single‐blind method) AND NOT (Ct animal AND NOT (Ct human and Ct animal)) OR (Pt clinical trial OR (Tw clin$ AND (Tw trial$ OR Tw ensa$ OR Tw estud$ OR Tw experim$ OR Tw investiga$)) OR ((Tw singl$ OR Tw simple$ OR Tw doubl$ OR Tw doble$ OR Tw duplo$ OR Tw trebl$ OR Tw trip$) AND (Tw blind$ OR Tw cego$ OR Tw ciego$ OR Tw mask$ OR Tw mascar$)) OR Mh placebos OR Tw placebo$ OR (Tw random$ OR Tw randon$ OR Tw casual$ OR Tw acaso$ OR Tw azar OR Tw aleator$) OR Mh research design) AND NOT (Ct animal AND NOT (Ct human and Ct animal)) OR (Ct comparative study OR Mh follow‐up studies OR Mh prospective studies OR Tw control$ OR Tw prospectiv$ OR Tw volunt$ OR Tw volunteer$) ANDNOT (Ct animal ANDNOT (Ct human and Ct animal)))

Appendix 7. ClinicalTrials.gov search strategy

Advanced search

Condition or disease: herpes zoster OR postherpetic neuralgia OR shingle OR PHN

Study type: Interventional Studies (Clinical Trials)

Intervention: adrenal cortex hormone OR glucocorticoid OR corticosteroid OR steroid OR prednisolone OR triamcinalone OR dexamethasone OR triamcinolone OR hydrocortisone OR prednisone

Appendix 8. WHO International Clinical Trials Registry Portal (ICTRP) search strategy

Advanced Search

(herpes zoster OR postherpetic neuralgia OR shingle OR PHN) in the Title

Recruitment status is ALL

Data and analyses

Comparison 1. Corticosteroids versus placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 The presence of PHN six months after the onset of acute herpetic rash	2	114	Risk Ratio (M‐H, Fixed, 95% CI)	0.95 [0.45, 1.99]
1.2 The main effect of prednisone compared with no prednisone on six months evaluation of pain (generic inverse variance)	1		RR Ratios (IV, Fixed, 95% CI)	Totals not selected
1.3 Serious adverse events	5	755	Risk Ratio (M‐H, Fixed, 95% CI)	1.65 [0.51, 5.29]
1.4 Non‐serious adverse events	5	755	Risk Ratio (M‐H, Fixed, 95% CI)	1.30 [0.90, 1.87]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Clemmensen 1984.

Study characteristics
Methods	Single centre RCT, double‐blind, placebo‐controlled, parallel‐group study (methods not described)
Participants	60 participants (33 males and 22 females) within 7 days of onset of HZ, 5 participants dropped out Age range: 16 to 86 years, 33 were 55 years or older. Among them, 20 participants received intramuscular ACTH (Synacthen depot). Exclusion criteria: (1) duration of symptoms (pain and/or cutaneous signs) beyond 7 days; (2) age under 16 years; (3) generalised HZ (more than 50 vesicles outside the affected dermatome); (4) history of, or current, malignant disease; (5) treatment with cytostatics and corticosteroids; (6) history or findings of peptic ulcer, psychosis, cardiac decompensation, hypertension, diabetes mellitus, adrenocortical disease or with symptoms of osteoporosis; (7) pregnancy.
Interventions	ACTH: ACTH depot 1 mg intramuscularly 3 times a week (Monday, Wednesday, Friday) amounting to a total of 7 injections Prednisone: orally, 45 mg daily during the 1st week, 30 mg daily during the 2nd week, and 15 mg daily tapered to zero during the 3rd week Comparison treatment: placebo
Outcomes	Primary outcome (mean duration of PHN: 4.2 months, which ranged from 1.5 to 10 months) and secondary outcomes (serious adverse events and non‐serious adverse events) were available.
Funding	None reported
Conflicts of interest	None reported
Notes	Conducted in Denmark We contacted the contact author to request missing information but received no answer.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	The method of randomisation was not described.
Allocation concealment (selection bias)	Unclear risk	The method of allocation concealment was not described.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	It was stated that a "double‐dummy" administration technique was used: matched oral and/or parenteral placebo was given to each participant. Placebo tablets or injections were indistinguishable from the active medication.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"Double‐dummy" administration technique
Incomplete outcome data (attrition bias) All outcomes	Low risk	5 participants dropped out, 1 in the prednisone group (because of increasing blood sugar), 1 in the placebo group (discontinued by the participant without specific reason), 3 in the ACTH group (2 participants developed uncomfortable dizziness, and 1 participant developed moderate periorbital oedema).
Selective reporting (reporting bias)	Low risk	Outcomes listed in the methods section were all reported.
Other bias	Low risk	No other potential bias was found.

Eaglstein 1970.

Study characteristics
Methods	Single centre, randomised, double‐blind, placebo‐controlled parallel design
Participants	35 participants with early, severely painful HZ admitted to the dermatology inpatient service Age: from 21 to 91 years, 24 of them older than 59 years of age. One participant dropped out. Exclusion criteria: hypertension, tuberculosis, lymphoma, leukaemia, bleeding peptic ulcers, diabetes, cardiac disease, or bacterial infections
Interventions	Unmarked red capsules containing either 8 mg of triamcinolone or lactose, 2 capsules 3 times daily (48 mg/day) for 7 days, and 1 capsule 3 times daily (24 mg/day) for 7 days, and 1 capsule twice daily (16 mg/day) for 7 days
Outcomes	Primary outcome (the presence of PHN six months after the onset of the acute herpetic rash) and secondary outcomes (serious adverse events and non‐serious adverse events) were available.
Funding	This trial was supported in part by Public Health Service dermatology training grant AM‐05262 from the National Institute of Arthritis and Metabolic Diseases, and by the Dermatology Foundation of Miami.
Conflicts of interest	None reported
Notes	Conducted in Miami, USA dermatology inpatient service We contacted the contact author to request missing information but received no answer.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random numbers were used; each participant was assigned a different code number and distributed to a group randomly.
Allocation concealment (selection bias)	Low risk	The code for each participant was opened after a 3‐year follow‐up and evaluation.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants were treated with unmarked red capsules prepared by the hospital pharmacy.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Unmarked red capsules were prepared by the hospital pharmacy. Comment: a double‐blind method was probably used
Incomplete outcome data (attrition bias) All outcomes	Low risk	Only one participant in the control group dropped out after five days because of a sudden increase in her BP.
Selective reporting (reporting bias)	Low risk	Outcomes listed in the methods section were all reported.
Other bias	Low risk	No other potential bias was found.

Esmann 1987.

Study characteristics
Methods	Multicentre randomised, double‐blind, placebo‐controlled parallel design. Number of losses to follow up: all participants were evaluated at week 26 except for one from the prednisolone group, who was last seen at week 10. She had not had pain since day 5.
Participants	84 participants (25 males and 53 females) within 4 days of onset of HZ Age at least 60 years, mean age: intervention group 72.8 years (SD 7.5); control group 71.4 (SD 8.1). Exclusion criteria: immunocompromise, pituitary or adrenal dysfunction, diastolic BP > 105 mmHg on entry day, signs of cardiac insufficiency, insulin dependent diabetes, bacterial infections, bleeding peptic ulcers, severe mental confusion, serum creatinine 150 mmol, receiving corticosteroid treatment
Interventions	800 mg aciclovir orally 5 times daily for 7 days and coded tablets containing either prednisolone or calcium lactate for 21 days. The dose of prednisolone was 40 mg daily for 7 days, 30 mg for 4 days, 20 mg for 3 days, 10 mg for 4 days, and finally, 5 mg for 3 days
Outcomes	The primary outcome (the presence of PHN six months after the onset of the acute herpetic rash) and secondary outcomes (serious adverse events and non‐serious adverse events) were available.
Funding	None reported
Conflicts of interest	None reported
Notes	Conducted in Aarhus and Copenhagen, Denmark We contacted the contact author to request missing information but received no answer.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	The method of randomisation was not described.
Allocation concealment (selection bias)	Unclear risk	The method of allocation concealment was not described.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	The trial report stated that a double‐blind method was used, and all participants were given coded tablets containing either prednisolone or calcium lactate.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	A double‐blind method was used.
Incomplete outcome data (attrition bias) All outcomes	High risk	Six participants were withdrawn, either because the inclusion criteria could not be upheld upon subsequent scrutiny or because of lack of compliance during the first 1 to 2 weeks. One of the six participants dropped out because of a possible side effect of prednisolone, and she could be included in the intention‐to‐treat analysis. But whether the lack of compliance of the other participants was due to inefficacy or side effects, and to which group those five participants were first assigned, were not specified.
Selective reporting (reporting bias)	Low risk	Outcomes listed in the methods section were all reported.
Other bias	Low risk	No other potential bias was found.

Whitley 1996.

Study characteristics
Methods	Multicentre randomised, double‐blind, placebo‐controlled parallel study with a 2 x 2 factorial design
Participants	208 immunocompetent participants older than 50 years of age who had localised herpes zoster that developed less than 72 hours before study enrolment. Five randomly assigned participants were not included in this analysis because they never received study medication; no case record forms were submitted. Two other participants were proven to have herpes simplex virus infection and were not included in the analysis. Of the 201 participants included in the analysis, 51 received aciclovir plus prednisone (24 males and 27 females, mean age 63), 48 received aciclovir plus prednisone placebo (21 males and 27 females, mean age 62), 50 received prednisone plus aciclovir placebo (26 males and 24 females, mean age 60), and 52 received aciclovir and prednisone placebo (25 males and 27 females, mean age 61). 32 participants were lost to follow‐up. Exclusion criteria: immunosuppressive therapy; cancer; women capable of conceiving and bearing a child; history of hypertension (diastolic BP >100 mmHg) or receiving antihypertensive therapy; osteoporosis or insulin‐dependent diabetes mellitus; receipt of other antiviral drugs or immunoglobulin products in the 4 weeks before the study began; history of glycosuria or hyperglycaemia
Interventions	Prednisone or a matched placebo orally 60 mg/d for days 1 to 7, 30 mg/d for days 8 to 14, and 15 mg/d for days 15 to 21 Aciclovir or a matched placebo orally as 800 mg 5 x daily, for 21 days. The four treatments regimens were aciclovir plus prednisone, aciclovir plus prednisone placebo, prednisone plus aciclovir placebo, and placebos for both aciclovir and prednisone.
Outcomes	The primary outcome (six‐month evaluation of pain (time to cessation of zoster‐associated pain) and secondary outcomes (one‐month evaluation of quality of life; serious adverse events and non‐serious adverse events) were available.
Funding	None reported
Conflicts of interest	None reported
Notes	Conducted in 15 university hospitals or affiliated clinics in USA Participants discontinued therapy because of influenza, conjunctivitis or iritis, nausea and vomiting, complete resolution of disease, cutaneous dissemination, hyperglycaemia, and bacterial pneumonia. We contacted the contact author to request missing information but received no answer.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computer‐generated randomisation code randomised participants to one of the four treatment groups
Allocation concealment (selection bias)	Unclear risk	Method of allocation concealment was not described
Blinding of participants and personnel (performance bias) All outcomes	Low risk	All research personnel remained blinded to drug assignment until the study was completed, and the database was locked. All matched medications were identical in taste and appearance.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All research personnel remained blinded to drug assignment until the study was completed, and the database was locked.
Incomplete outcome data (attrition bias) All outcomes	Low risk	32 participants dropped out: 7/51 in the aciclovir plus prednisone group, 6/48 in the aciclovir plus prednisone placebo group, 9/50 in the prednisone plus aciclovir placebo group, and 10/52 in the wholly placebo‐treated group. Missing data were equal among the treatment groups, and an intention‐to‐treat analysis was performed.
Selective reporting (reporting bias)	Low risk	Outcomes listed in the methods section were all reported
Other bias	Low risk	No other potential bias was found.

Wood 1994a.

Study characteristics
Methods	Multicentre randomised, double‐blind, placebo‐controlled parallel study.
Participants	Adults over 18 years of age without immune dysfunction due to cancer or immunosuppressive therapy, who presented with a clinical diagnosis of HZ, confirmed by one of the investigators, and had a rash for 72 hours or less and at least moderate pain. 400 participants enrolled and assigned to 4 groups: aciclovir for 7 days with corticosteroids (99 participants, 37 males and 62 females, mean age 59) aciclovir for 7 days without corticosteroids (101 participants, 39 males and 62 females, mean age 58) aciclovir for 21 days with corticosteroid (99 participants, 39 males and 60 females, mean age 60) aciclovir for 21 days without corticosteroid (101 participants, 38 males and 63 females, mean age 59) 51 participants were withdrawn. Losses to follow up: 2 participants in 7‐day aciclovir with corticosteroids; 3 participants in 7‐day aciclovir without corticosteroids; 2 participants in 21‐day aciclovir with corticosteroids; 3 participants in 21‐day aciclovir without corticosteroids Exclusion criteria: pregnant women and women of childbearing potential who were not adequately protected by contraception; renal insufficiency (serum creatinine concentration, more than 1.8 mg per dL), hypertension (diastolic BP >110 mmHg), insulin‐dependent diabetes, or random blood glucose > 216 mg/dL (12 mmol/L); history of peptic ulceration, severe psoriasis, or hypersensitivity to aciclovir; and people receiving barbiturates, anticonvulsant drugs, systemic corticosteroids, rifampicin, or specific antiviral therapy for the present infection
Interventions	Aciclovir (800 mg orally) five times daily, beginning on day 0. The participants in the groups assigned to 7 days of aciclovir therapy (with or without corticosteroid) received matching placebo beginning on day 7. Prednisolone (5 mg tablets) according to the following schedule: on days 0 to 6, 40 mg per day; days 7 to 10, 30 mg per day; days 11 to 14, 20 mg per day; days 15 to 18,10 mg per day; and days 19 to 21, 5 mg per day (total dose 535 mg)
Outcomes	The primary outcome (up to month six to assess PHN, and time to complete cessation of pain) and secondary outcomes (serious adverse events and non‐serious adverse events) were available.
Funding	Supported by a grant from the Wellcome Research Laboratories, Beckenham, United Kingdom
Conflicts of interest	None reported
Notes	Conducted in 4 clinical centres in the United Kingdom Disclosures of interest: none reported We contacted the contact author to request missing information but received no answer.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computer‐generated randomisation code was used.
Allocation concealment (selection bias)	Low risk	The randomisation code was stratified by study centre to assigned participants in blocks of eight to either group; it indicated that allocation concealment might be performed.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	It was stated that it was a double‐blind study, and the participants in the groups not receiving corticosteroid received matching placebo tablets.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Double‐blind method was used.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Withdrawals of participants and the reasons were balanced equally across all groups. 51 participants were withdrawn: 14 in aciclovir for 7 days with corticosteroid group, 10 in aciclovir for 7 days without corticosteroid group, 13 in aciclovir for 21 days with corticosteroid group, 14 in aciclovir for 21 days without corticosteroid group. Reasons were deviation from protocol, adverse events, loss to follow‐up, no reason given, and death.
Selective reporting (reporting bias)	Low risk	Outcomes listed in the methods section were all reported.
Other bias	Low risk	No other potential bias was found.

ACTH: adrenocorticotrophic hormone
BP: blood pressure
HZ: herpes zoster
PHN: postherpetic neuralgia
RCT: randomised controlled trial
SD: standard deviation

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Agarwal 1991	Confirmed not a true RCT.
Brusco 1993	Confirmed not a true RCT.
Chang 2004	PHN was defined as pain persisting at the site of shingles 2 weeks after the onset of acute rash.
Cui 2002	Study author confirmed it was not truly randomised.
Cui 2017	The intervention group was treated with steroids plus anaesthetics.
Dureja 2010	Methylprednisolone was given by epidural injection.
Guo 2001	The routine treatments were mismatched between 2 groups.
Hao 2002	Study author confirmed this was not truly randomised.
Huang 2004	Study author confirmed this was not truly randomised.
Huang 2010	Confirmed not a true RCT.
Ji 2009	The intervention group was treated with methylprednisolone plus bupivacaine.
Jiang 2005	The routine treatments were mismatched between 2 groups.
Jiang 2008	Confirmed not a true RCT, and the duration of herpes zoster was not specified.
Keczkes 1980	The control group was treated with carbamazepine.
Li 2000	Confirmed not a true RCT.
Li 2002	Study author confirmed this was not truly randomised.
Liao 2005	PHN was defined as pain persisting 1 week after total decrustation.
Lin 2002	This trial exceeded our 7 days to treatment criterion.
Lin 2005	Study author confirmed this was not truly randomised.
Liu 2003	PHN was defined as pain persisting at the site of shingles 2 weeks after the onset of acute rash.
Liu 2005	PHN was defined as pain persisting at the site of shingles 2 weeks after the onset of acute rash.
Ma 2000	Confirmed not a true RCT.
Ma 2002	Study author confirmed this was not truly randomised.
Makharita 2012	The intervention group was treated with dexamethasone plus bupivacaine.
NCT05208918	Steroid was given by transforaminal epidural injection.
Ni 2017	The intervention group was treated with triamcinolone plus lidocaine.
Pasqualucci 2000	The intervention group was treated with methylprednisolone plus bupivacaine by epidural injection.
Rijsdijk 2013	Methylprednisolone was given by intrathecal injection.
Shi 2008	Confirmed not a true RCT.
Song 2009	Confirmed not a true RCT.
Tang 2004	Study author confirmed this was not truly randomised.
van Wijck 2006	The intervention group was treated with methylprednisolone plus bupivacaine by epidural injection.
Wang 2004	This trial exceeded our 7 days to treatment criterion.
Xu 1999	This trial exceeded our 7 days to treatment criterion.
Yang 2000	The routine treatments were mismatched between 2 groups.
Yang 2002	PHN was defined as pain persisting at the site of shingles 2 weeks after the onset of acute rash.
Yang 2010	Confirmed not a true RCT, and PHN was not clearly defined. The duration of herpes in some participants was longer than 7 days after the rash onset.
Yin 2004	The routine treatments were mismatched between 2 groups.
Yin 2005	The routine treatments were mismatched between 2 groups.
Zeng 2011	Confirmed not a true RCT.
Zhang 2003	Study author confirmed this was not truly randomised.
Zhang 2004	The routine treatments were mismatched between 2 groups.
Zhang 2005	This trial exceeded our 7 days to treatment criterion.
Zheng 2004	The routine treatments were mismatched between 2 groups.
Zhou 2000	Study author confirmed this was not truly randomised.
Zhou 2008	Not a true RCT.

PHN: postherpetic neuralgia
RCT: randomised controlled trial

Characteristics of studies awaiting classification [ordered by study ID]

Hu 2001.

Methods	A single‐centre, randomised controlled parallel trial
Participants	45 participants with confirmed herpes zoster and without contraindications to glucocorticosteroids were enroled. They were randomly assigned to the aciclovir plus dexamethasone group (20 participants) or the aciclovir alone group (25 participants).
Interventions	In the aciclovir plus dexamethasone group, participants were given 0.25 g aciclovir intravenously every 8 hours for a total of 10 days, and 5 mg dexamethasone intravenously daily. In the control group, aciclovir alone was administered in the same way as in the other group.
Outcomes	Time to cessation of new herpes eruption, time to crusting, and the presence of PHN at 1 month after the rash onset
Notes	The exact course of disease from onset of herpes zoster to receipt of treatment was not specified. We tried to contact the author, but have not yet received a reply.

Differences between protocol and review

In the 2010 update, we assessed risk of bias using new methods and added a summary of findings table (Chen 2010). We introduced a change from the protocol to exclude quasi‐RCTs (He 2006), because RCTs are thought to be the only way to prevent systematic differences in baseline characteristics between intervention groups, according to the latest Cochrane Handbook for Systematic Reviews of Interventions. However, we found no quasi‐RCTs.

In the 2012 update (Han 2013), we split 'blinding' into assessments of blinding of participants and personnel (performance bias) and blinding of outcome assessors (assessment bias), and revised the wording of assessments to high, low, or unclear (Higgins 2017). We added information about the summary of findings table to the methods section.

In the 2022 update, we made changes to the methods according to MECIR (Higgins 2023a). There were also changes to the team of authors. See Other published versions of this review.

Contributions of authors

Xin Jiang: literature search, study selection, risk of bias assessment, data extraction, data analysis, writing review, updating review.
Yanbo Li: literature search, study selection, data extraction, writing review.
Ning Chen: literature search, data extraction, updating review.
Muke Zhou: literature search, risk of bias assessment, updating review.
Li He: conceived and designed study, literature search, study selection, risk of bias assessment, data extraction, data analysis, writing review, updating review.
All authors reviewed and approved the review.

Sources of support

Internal sources

• No sources of support supplied, Other

  No sources of support supplied

External sources

• No sources of support supplied, Other

  No sources of support supplied

Declarations of interest

Xin Jiang: none known
Yanbo Li: none known
Ning Chen: none known
Muke Zhou: none known
Li He: none known

New search for studies and content updated (conclusions changed)