As Lord Justice Jackson observed in his preliminary report in May 2009, the vast numbers of electronic documents in substantial litigation present an acute dilemma for the justice system (www.practicallaw.com/5-385-9487). E-disclosure can be of great assistance to the court but, at the same time, it risks being very expensive. In large e-disclosure exercises, there is also a risk of inadvertently disclosing privileged documents, particularly if the document review is not subject to quality control.

The Jackson reforms sought to mitigate costs by introducing more flexibility as to the scope of search and a timetable for discussions between the parties about e-disclosure (see feature article "Jackson reforms: what commercial parties need to know"). However, so far, the reforms do not appear to have significantly reduced disclosure costs.

Since April 2013, Civil Procedure Rule (CPR) 31.5(7) has replaced the presumption in favour of standard disclosure with a menu of disclosure options for multi-track cases (the menu approach). Standard disclosure is no longer the default position. Instead, the court must decide on the order that is most appropriate, having regard to the overriding objective and the need to limit disclosure to documents that are necessary to deal with the case justly. Potential orders include:

As a further change to the rules, CPR 31.5(5) provides that, two weeks before the first case management conference (CMC), the parties must file and serve a disclosure report verified by a statement of truth. The disclosure report should briefly describe what documents exist, where they are located, how they are stored, and the likely costs of giving standard disclosure. If the parties have already exchanged electronic documents questionnaires (EDQ), each party should append its EDQ to its disclosure report. The EDQ is a form containing information about a party's electronic documents and its proposals for a reasonable and proportionate search.

At least seven days before the first CMC, the parties must discuss and seek to agree their proposals in relation to disclosure (CPR 31.5(5)). If they agree proposals, the court can approve them without a hearing (CPR 31.5(6)). Failing this, CPR 31.5 envisages the court making an order tailored to the circumstances of the case at the first CMC.

The key advantage of the menu approach is in ensuring that the court and the parties focus on the extent of the disclosure required at an early stage. It also allows the court the flexibility to select an order that is appropriate to the case, rather than simply defaulting to standard disclosure.

However, standard disclosure continues to be routinely ordered and, even where another option from the menu is selected, it does not necessarily follow that significant savings will be made. Indeed, very large volumes of documents may still have to be reviewed and disclosed; for example, if disclosure is ordered on an issue-by-issue basis, this may help to focus the disclosure but will not necessarily limit the number of documents reviewed. A wide-ranging review may also be necessary to identify the documents that a party wishes to rely on.

So, while the Jackson reforms to disclosure continue to take root, how else can parties seek to control the costs of the process? The answer is that law firms are adopting increasingly sophisticated strategies to tackle the e-disclosure challenge, including alternative ways of structuring document review exercises and capitalising on advances in review technologies.

The use of date range limitations, custodian selection, keyword searching and de-duplication has become standard practice for many large e-disclosure exercises. However, other, more advanced technologies are now beginning to take hold; for example, early case evaluation technology and predictive coding.

To date, predictive coding has been used relatively infrequently in English litigation but is more common in the US. In simple terms, predictive coding is an e-disclosure software tool that allows the review team to prioritise documents that are more likely to be relevant to the disputed issues and so reduces the overall number of documents that need to be reviewed. It can be used to allocate documents for first pass review or within a quality control exercise.

But this is not to say that it is a one-click or automated solution, nor does it replace the need for individuals to review documents and make decisions about relevance and privilege. Running a predictive coding exercise also requires sophisticated e-disclosure technology, expertise and a good understanding of statistics. Another reason that predictive coding is not yet commonly used in England is the absence of judicial endorsement in case law (see Know how article "Judicial approval for predictive coding: canvassing opinion").

While predictive coding is not a cure-all for e-disclosure, it can reduce the time and costs of the process in appropriate cases and when used correctly. To this extent, its use can also arguably be consistent with Practice Direction 31.B.9, which requires the parties to discuss the use of tools and techniques to reduce the burden and cost of e-disclosure, including the use of software tools. So, it seems likely to become a more familiar part of the disclosure landscape in the coming years.

Adopting a less centralised approach to structuring and staffing a document review exercise can also result in significant savings. This may be achieved through legal process outsourcing, where a third party conducts the initial review of documents, or through law firms employing staff in less expensive locations, whether onshore or offshore (see Know how articles "Legal process outsourcing: opportunities and risks"; "Half a world away: outsourcing document review").

This will help to reduce the average hourly rates of the individuals reviewing the documents. The challenge of these strategies is in ensuring that the process works, maintains flexibility as the issues develop and can be justified to the court.

Herbert Smith Freehills LLP has developed a structured approach to e-disclosure. A key aspect has been opening an office in Belfast and establishing a large team of solicitors and law graduates to undertake legal process work, including intensive document review (see Know how article "Herbert Smith's Belfast operation: keeping document review close to home"). This has substantially reduced the cost of reviewing documents without losing any control over the process.

In addition, the firm has invested in new e-disclosure technology, Recommind Axcelerate, for use by the Belfast disclosure teams and within the firm's network to make the process as efficient as possible. To bridge the gap between the law, practice and technology, the firm also established a core disputes data management team of e-disclosure specialists, including several qualified lawyers. Apart from the costs savings, this structured approach allows for the consistent application of quality control and better knowledge sharing between the reviewers and the legal team.

Maintaining quality control is a key factor to take into account when considering which strategy to adopt for dealing with e-disclosure. In any large disclosure review there is a risk that documents containing legally privileged information will be inadvertently offered for inspection.

The recent Court of Appeal decision in Rawlinson and Hunter has emphasised that the test is whether it is obvious that the privileged document was disclosed by mistake ([2014] EWCA Civ 1129, www.practicallaw.com/8-581-8266). If there is no obvious mistake, then privilege has been waived (for background, see feature article "Waiver of privilege: all is not lost"). This highlights the importance of reviewers having a clear understanding of the rules of legal professional privilege and their application to the case, and of ongoing quality control and privilege testing during the review.

Another practical implication, particularly for large scale e-disclosure exercises, may be increased engagement between parties before disclosure about how to deal with inadvertently disclosed documents. In the US, clawback agreements, in which the parties set out the terms on which they will return privileged documents, are often used.

The status and enforceability of these agreements is uncertain under English law, but there is scope for argument that CPR 31.B(3)(f) envisages this type of arrangement. It provides that, where appropriate, the parties should discuss and agree the methods to be used to identify privileged documents and other non-discloseable documents, and for dealing with privileged or other documents that have been inadvertently disclosed. The Technology and Construction Court's e-disclosure protocol also provides for the use of clawback agreements. Given the timely reminder of Rawlinson and following the US's example, these agreements may become more widely used than at present, in which case it may not be long before their validity is tested.