South African National AIDS Helpline, The

This study offers a quantitative analysis of calls made to the national AIDS Helpline in South Africa over a three year period. The Helpline has operated since 1992, providing information, counselling and referrals to callers in all 11 South African languages. Analysis of Helpline data capture forms found that the overall volume of calls to the Helpline has declined over time, but that a growing proportion of the calls are for counselling, rather than basic information.


Seventy-five percent of callers are under 30 and a growing proportion of callers are disclosing their HIV status. Analysis of Helpline call and caller trends can be utilised to inform HIV/AIDS communication policy.

Between May 2001 and December 2003, a total of 5,351,086 calls were registered on the system. For technical reasons, not all calls were registered during some periods and it is estimated that the actual number of received calls is probably closer to 7,000,000. Based on actual calls registered, the Helpline received an average of 201,476 calls per month, although average calls per month ranged from 98,307 to 392,667 (Figure 2). Calls were categorised into one of four groups:

• Lost calls
• Unlogged calls
• Hoax calls
• Genuine calls

The distribution of calls, by type, over time is shown in Figure 3. Lost calls are defined as calls that are not answered. Between May 2001 and December 2003, 22.8% of calls were lost. The proportion of lost calls has decreased considerably over time, from 33% during May-December 2001 to 9% during May-December 2003.

Unlogged calls refer to calls that are answered, but not captured on the call sheets. Approximately one-fifth of the calls (20.5%) are not logged into the system. There are three possible explanations for this. First, calls with a duration of less than one minute are not logged. Second, answered calls are sometimes lost when being transferred from the frontline operators (who filter calls) to the counsellors. Third, counsellors do not always log every call, either due to tardiness or because the time between incoming calls is too short to allow for the datasheets to be completed. The proportion of answered, unlogged callsincreased sharply from May-December 2001 (11%) to May- December 2002 (31%), but then decreased again to 25% in May-December 2003 (Figure 3). While it is unclear why there was a sharp increase, the decrease may be a result of an increase in the number of calls for counselling, which are longer in duration than those requesting information alone. The longer call duration increases the chance that the call will be logged into the system.

The third category refers to hoax calls. Of all captured calls, 86% were classified as hoax calls. A sharp increase in hoax calls as a proportion of overall calls was seen between May-December 2002 (45%) and May-December 2003 (60%) (Figure 3).
Hoax calls are typical amongst tollfree lines in South Africa. The nature of hoax calls varies, but includes silent callers, children playing on the line, abusive callers and wrong numbers. Such calls are demotivating to counsellors, waste time, and are costly overall. Possible approaches to diminishing hoax calls have been explored (UNAIDS 2002) and include the possibility of classifying hoax calls by category to allow for a better understanding of the distribution of types of hoax calls, and using such data to develop a response strategy. Another option is to employ a Caller Line Identity system to assess the provenance of calls and to block calls repeatedly made from the same number.

The fourth category includes genuine calls – defined as answered calls in which information, referral and/or counselling is provided to the caller. Figure 4 shows the number of genuine calls received by the Helpline between July 2000 and October 2003. The number of genuine calls fell from 20,689 during July 2000 to 12,329 in April 2001.1 The April 2001 call rate had more than doubled by August 2001, reaching an all-time peak of 31,020 calls. This increase was followed by a decrease to 8,690 calls in June 2002. Between July 2002 and April 2003, the number of calls fluctuated between 10,107 and 15,242. Thereafter the number of calls declined, reaching the lowest recorded level for the whole review period during the last month under study – December 2003 – with 6,543 calls. As Figure 3 shows, the percentage of genuine calls among overall calls decreased from 8% in May-December 2001 to 6% in May-December 2002, after which it stabilised.