Telephone surveys have played a vital role in public health research and practice for many decades [1, 2]. Telephone interviews have a number of important advantages over face-to-face surveys: (1) they are usually more cost and time effective; (2) the sampling frame can include areas that may be hard for interviewers to access in person; and (3) respondents have a sense of anonymity once they understand that their telephone number was chosen at random, which can increase the validity of information being reported . In recent years, several knowledge, attitudes and beliefs (KAB) telephone surveys have been conducted successfully in Ireland, for example, the Sexual Abuse and Violence in Ireland Study (SAVI-2001), the Irish Contraception and Crisis Pregnancy Study (ICCP-2003)  and the Irish Study of Sexual Health and Relationships (ISSHR-2006) .
Traditionally, sampling frames for telephone public health surveys in Ireland and elsewhere have been limited to landline telephone numbers . Recent research in the US, Australia, and Europe has demonstrated that the percentage of the population who do not have a landline telephone in their home has steadily decreased in the last few decades [1, 8–11]. For example, it is estimated that approximately 23% of US households in 2009 were mobile telephone only households, an increase from 2% in 2003 . In Ireland, a recent national face-to-face household survey of lifestyles, attitudes and nutrition (SLÁN 2007)  revealed that 61% of respondents had access to a mobile and landline telephone where they lived; however, a substantial percentage (23%) of individuals only had access to a mobile telephone . Extensive research in the US has revealed that households using mobile telephones exclusively are disproportionally younger, male, single, and living in rental accommodation when compared to households that have a landline telephone [7, 14]. This population of adults leading a 'wireless only' lifestyle  are also more likely to have higher rates of detrimental health behaviours compared to adults who live in households that use a landline telephone .
Non-coverage of households without a landline telephone is a major concern of telephone survey researchers [7, 9]. It is clear that sampling mobile phone users in national surveys is vital to gain access to the growing proportion of households that use mobile telephones extensively or exclusively . Recently, the feasibility of recruiting respondents via mobile telephones for national health surveys, such as the 2007 California Health Interview Survey  and the Behavioural Risk Factor Surveillance System (BRFSS) , has been investigated. As a result of these and other methodologically pioneering studies, the challenges associated with conducting research using mobile telephones have been debated within the literature [1, 17]. These challenges include, but are not limited to, the following: (1) safety: the respondent could receive a mobile telephone call in any location (e.g. whilst driving a car) (2) privacy: the respondent could be responding to the survey in a location where they may not be able to comfortably disclose sensitive information or they could feel that an unsolicited call on their personal mobile telephone is an invasion of their privacy, unlike the typical private household landline situation of traditional surveys; (3) eligibility: it can be difficult to determine whether the mobile phone owner is an adult and/or eligible for a study; (4) cost for the respondent: the respondent may incur a charge if they are accessing a mobile network other than their service provide ('roaming') or on a pre-paid plan that charges for incoming calls (often referred to as 'Receiver Party Pays'- commonly used in countries such as the US, Canada, and China) ; (5) cost for the commissioning agent: the cost of calling mobile telephones to recruit respondents is more expensive compared to landline telephones; (6) legality: calling private mobile telephones may breach communication legislation (e.g. the US 1991 Telephone Consumer Protection Act); (7) identification: it is not as easy to identify whether the mobile phone is being used for business or personal matters; (8) connections problems: once a connection has been established, it is possible for the call to be 'dropped' due to lack of mobile network coverage or battery supply; (9) respondent burden: given that people are often under special time constraints and special pressures when speaking on mobile telephones, surveys administered via a mobile telephone may need to be shorter than those conducted via landline telephones ; and (10) singular nature of mobile telephone: a landline telephone is normally used by several members of a household, whereas a mobile telephone has generally one primary user. In surveys, if a respondent answers a landline telephone and is either ineligible or unwilling to complete a survey, an interviewer can prompt for another eligible individual in the household to complete the survey; this is not possible using mobile phones.
All of the issues outlined above have the potential to seriously impede on the success of surveys that utilize mobile telephones to recruit respondents. For example, it has been suggested that response rates to telephone surveys in Ireland, the UK, and elsewhere have generally diminished over recent years because householders are often inundated by requests from market research companies (MRC) to participate in telephone surveys . Using mobile telephones to recruit respondents has considerably reduced the response rate of national surveys in the US . It is vital, therefore, that considerable efforts are made to try and overcome these obstacles.
The aim of this paper was to document the challenges encountered during a recent Irish KAB survey called the 2010 Irish Contraception and Crisis Pregnancy Survey (ICCP-2010). ICCP-2010 is a cross-sectional, nationally representative telephone survey of knowledge, attitudes and behaviours in relation to crisis pregnancy, contraception and sexual health, among adults aged 18-45 years living in Ireland (n = 3002). The survey is unique for two major reasons: (1) it was the first national survey in Ireland to split the recruitment for the survey over mobile and landline telephones, which enhances the representativeness of the sample; and (2) respondents were not incentivised to participate but a high overall response rate of 69% was achieved, which is unusual for surveys using mobile telephone recruitment . The aim of this paper is to outline how the ICCP-2010 was conducted, with a particular focus on describing the main challenges encountered while recruiting a large random sample using mobile telephones and the outlining the practical steps taken to overcome these challenges successfully to achieve a high response rate from a nationally representative sample of the general population.