Time course of vasovagal syncope with whole blood donation


  • 3D-S08-03

Peter Tomasulo, Blood Systems, Inc. Scottsdale, Arizona, USA
E-mail: ptomasulo@bloodsystems.org


Background  Improving the safety of the donation experience will reduce donor injuries and increase donations, donation frequency and donor satisfaction. Understanding the physiology of donor reactions supports selection of effective interventions to reduce risk.

Aims  The examination of the time course of donor vasovagal syncopal reactions (VVS) to determine when most reactions occur supports the development of appropriate theories about cause of the majority of the reactions and the application of interventions to reduce the risk of a significant number of reactions.

Materials and Methods  A database of more than 900 000 donor registrations and more than 500 000 whole blood donation events was examined for the time course of VVS reactions by reviewing the onset time of the reactions. The donor experience was divided into three periods based on the theory that the risk of reaction depends on different factors in each of the periods. The time course was analyzed using frequency distribution counts stratified by donation status and separately by gender.

Results  There were 956 766 registrations, 554 513 attempted whole blood donations, 536 907 complete donations and 17 606 incomplete donations. The mean draw time of a complete donation was 8.21 (95% CI, 8.2–8.21) min and for an incomplete donation was 8.91 (95% CI, 8.8–9.01) min. The reaction rate for Period 1 (before venipuncture) was 0.045/1000 registrations; for Period 2 (venipuncture to 4 min post-venipuncture) and Period 3 (4 min post venipuncture to last reported VVS reaction -265 min), the reaction rate was 3.5/1000 for women and 1.5/1000 for men. There was a steady increase in the reaction rate during phlebotomy and the rate peaked at the time of needle removal. After needle removal there were peaks in the VVS rate at 5 and 9 min. The reaction rate declined steadily thereafter.

Discussion  The peak reaction rates occur within a time period of 10 min beginning 1 min before needle removal. The reactions which have the most significant risk of donor injury occur later than this and often occur after the donor leaves the donation site. Approximately 10% of VVS reactions in male donors and 25% of VVS reactions in female donors occur more than 15 min after the needle is removed. The latest VVS reaction was reported to have begun at about 4.5 h after needle withdrawal.

Conclusion  It is our thought that preventive interventions for reactions which occur during phlebotomy might be slightly different from the preventive measure aimed at reducing risk to donors after assuming upright posture. Once the donor stands up, relative hypovolemia plays a greater role and interventions should be aimed at preventing changes in blood pressure due to hypovolemia during this period.


The purpose of this study is to provide specific information about the time course of vasovagal syncope (VVS) in relation to whole blood donation from the time the donor registers at the blood donation site until the last reaction occurs. Understanding this pattern may shed light on the physiologic processes during and after donation in relation to syncope and injury. This study may also facilitate selection of interventions and assessment of the efficacy of interventions in future investigations.

Materials and methods

Data acquisition

We studied the reactions at 15 Blood Systems, Inc. (BSI) centres from January 1 to December 31, 2007 and provide a descriptive analysis of the operations of a large non-profit blood system producing more than 1·1 million red cell units annually. All BSI blood centres utilizing the MAK Progesa (MAK/Progesa, Paris, France) system use one standard operating procedure manual which controls the blood donation process. Donor reactions are identified, managed and documented according to that standard operating procedure. As was previously reported, an extract of donor and donation data for allogeneic whole blood donations was obtained from the blood establishment computer system by accessing the information through the blood bank Data Warehouse using Sagent Information Studio 5·5·1 (Sagent Technology, Inc, Pasadena, CA). Apheresis donations were excluded from this analysis. Donors gave 500 ml of blood plus an additional quantity (40–50 ml) which was used for donor testing. All donations were performed with a 16-gauge needle. The records capture donor demographics, biometric characteristics, clinical measurements ascertained before the donation, and donation characteristics. Donors for WB must weigh ≥50 kg, have a systolic BP between 90 and 180 mmHg, diastolic BP between 50 and 100 mmHg, pulse between 50 and 100 bpm, hematocrit ≥38% and must not have heart or lung disease, which prevents normal daily activity [1,2].

Time periods

For this study, we divided the donation sequence into three Periods. Period 1 begins when the donor registers to donate blood and ends at venipuncture. Period 2 begins at venipuncture and ends 4 minutes after the needle is withdrawn. Period 2 is ∼14 minutes in duration and was selected to include the phlebotomy and the time after phlebotomy during which the donor remains recumbent. Period 3 begins 5 minutes after needle withdrawal and is meant to begin at the time that many BSI donors change from recumbent to upright posture, and continues until the report of the last VVS reaction (261 minutes). The time at which the donor assumes upright posture was not recorded but was estimated to generally be <5 minutes by staff. Period 3 includes VVS reactions that occur after the donor leaves the donation site.

Adverse events

The adverse event studied was VVS with documented loss-of-consciousness (LOC). Adverse reactions are reported on standardized forms that include procedure end time (time needle removed from donor’s arm), reaction onset time, location where the adverse reaction occurred, specific symptoms, and duration of the reaction. The onset time recorded was the time of early symptoms. The actual time of the LOC was not recorded.

Donors call the centre to report reactions that occur off-site. Off-site reactions rarely are observed by blood centre staff. Information on off-site reactions is captured on the same form if reported by donors, and occasionally is accompanied by reports from physicians or hospitals. During this 2007 study period, donors were not encouraged to drink water before donations or to perform muscle tensing exercises.

Duration of phlebotomy

The time of venipuncture and needle withdrawal are recorded for each donation. The duration of phlebotomy was calculated by subtracting the venipuncture time from the needle withdrawal time.

Onset time of reaction

The onset time of VVS (in minutes) was calculated based on the recorded onset time of reaction in comparison with the phlebotomy end time documented for the donation event (T = 0). Records classified as “missing onset time” include records with time of more than 10 minutes before needle removal (15 records) and those without an onset time (34 records). Removing these records (3%) was not thought to affect the validity of the timeline analysis.

Definition of donations

Donations were defined based on phlebotomy status recorded on the donation record. Incomplete donations had <450 mls and complete donations had 450–550 mls.

Time course of vasovagal syncope

Time course of VVS was analyzed using frequency distribution counts stratified by donation status (complete or incomplete) and separately by gender (male or female). The overall rate of VVS in each point of time in relation to needle removal was calculated by dividing the number of VVS in each one minute segment by the total number of donations in the appropriate category. Cumulative data were calculated over the time course by comparing the number of VVS events up to each time point compared to the total number of events in that category. Draw time distribution, means and confidence intervals were also calculated for each stratified group of interest. Data cleaning and statistical analysis were performed using STATA 11 SE (STATA Corporation, College Station, TX).



During 2007, the centres registered 956 766 donations. Of the 956 766 registrations, 554 513 attempted to donate whole blood in a manual system; of these, 536 907 donated complete and 17 606 donated incomplete whole blood units. There were 227 345 male and 327 168 female complete donations. There were 5564 male and 12 042 female incomplete donations (Table 1).

Table 1.   Summary of donation status and characteristics - Gender, Draw time, Vasovagal syncope, Distribution and Time
 IncompleteCompleteComplete and incomplete donations
  1. VVS, vasovagal syncope.

Donations556412 04217 606221 781315 126536 907227 345327 168554 513
Percent of Category326834159974159100
Mean draw time
 Minutes(95%Cl) (records with 0–30 minute draw time)9·21 (9·03–9·4)8·76 (8·63–8·89)8·91 (8·8–S.01)7·66 (7·61–7·64)8·61 (8·60–8·63)8·21 (8·2–8·21)7·66 (7·65–7·68)8·62 (8·61–8·63)8·22 (8·22–8·24)
 Vasovagal Syncope (VVS)541061602871024131134111301471
 Rate/1000 donations9·78·89.11·33·22·41·53·52·7
Minutes to
 Last VVS6106106103265265103265265
 50% of VVS000587376
 95% of VVS485195142184840

Duration of whole blood donation

There were 554 481 donations with venipuncture and needle withdrawal time; 553 159 donations were in the duration range of 0–30 minutes (99·76% of draw times); 535 805 and 17 354 were complete and incomplete donations, respectively.

Of the records with duration, 138 had negative values, 1184 had times greater than 30 minutes (together 0·2%) and those were removed from the calculation.

Of the donations in the 0–30 minute range, 2·75% (15,225) had calculated phlebotomy duration of 1 to 4 minutes, 52% had phlebotomy duration of 5–8 minutes and 93% had duration of 5–15 minutes. The mean draw time for all donations is 8·23 (95% CI 8·22–8·24); 7·66 (95% CI 7·65–7·68) and 8·62 (95% CI 8·61–8·63) for donations by men and women respectively. Female donations required 8·62 minutes which is significantly different from the male time of 7·66 minutes (P < 0·0001). The mean draw time for a complete donation is 8·21 (95% CI 8·2–8·21), while the mean for an incomplete collection is 8·91 (95% CI 8·8–9·01). (Table 1).

Vasovagal syncope

During 2007, there were 1563 episodes of VVS; 43 episodes in Period 1 and 1520 episodes in Periods 2 and 3. Only 1471 of the 1520 (96·8%) of the post venipuncture VVS reactions had documented onset times. After venipuncture, women had 1130 VVS reactions for a rate of 3·5/1000, while men had 341 VVS reactions for a rate of 1·5/1000. Two percent of the VVS reactions associated with incomplete donations and 8·7% of the VVS reactions associated with complete donations occurred off-site.

Figure 1 displays the rate of reaction in one-minute periods calculated by dividing the number of VVS by 554,513 donations. The overall reaction rate increases during phlebotomy to 0·06 VVS/1000 donations one minute before needle withdrawal. Three peaks in rate are noted at t = 0, t = 5 and t = 9 (0·46/1000, 0·17/1000 and 0·15/1000 respectively). The rate then falls steadily with sporadic VVS reactions until 265 minutes after needle withdrawal. Many donors leave the donation site by 15 minutes after phlebotomy. There is no active off-site donor surveillance; thus, the prevalence of off-site syncope may be underestimated. Eighty per cent of reactions occurred within 15 minutes, 97% within 60 minutes and 99·5% within 120 minutes of needle withdrawal. The last reported reaction occurred at 265 minutes post needle withdrawal.

Figure 1.

 Overall vasovagal syncope rate in allogeneic WB intended donations, 2007.

During Period 1, there were 43 VVS reactions. We estimate that Period 1 is about 5–30 minutes, though the range is great depending on staffing and number of registrants. The rate of VVS reaction during this prevenipuncture period (Period 1) was extremely low with 0·045/1000 registrations.

The donors who gave the 536 907 complete whole blood donations had 1311 VVS reactions for a rate of 2·4 per 1000. Four hundred forty of the 1311 VVS reactions (34%) occurred during Period 2. During Period 2, the VVS rate is 0·002/1000 venipunctures at the beginning of phlebotomy and increases to 0·03/1000 during the period 1–2 minutes before needle removal. Then at t = 0, the VVS rate is 0·28/1000 (∼11% of the total number of VVS events occur during this 1 -minute time period.). Between needle removal and 4 minutes post, the VVS rate is 0·1–0·12/1000. A small peak in the reaction rate at 5 minutes (0·17/1000) begins Period 3. The rate then drops to a low of 0·09/1000 donations and then there is another peak at 9 minutes 0·16/1000 donations). There are sporadic reactions because of complete donations over the next 4 hours (Table 1, Figs 1–3).

Figure 2.

 Vasovagal syncope onset time distribution in complete and incomplete WB donations, 2007.

Figure 3.

 Frequency of vasovagal syncope reactions among allogeneic WB intended collections, 2007.

There were 17 606 incomplete whole blood donations. The donors who failed to give complete units had 160 reactions for a reaction rate of 9·1/1000 venipunctures. Of these VVS reactions, 146 (91%) occurred during Period 2. Thirty-one of these VVS reactions occurred during phlebotomy but before t = 0. One-hundred-four occurred at t = 0 and 11 occurred between t = 0 and t = 4. (Figs 2–3) Only 25 VVS (16% of the total) occurred after the needle was withdrawn. The rate of reactions in the group donating incomplete units began at 0·06/1000 (# VVS/17,606) and increased to 1/1000 in the minute before the needle was removed. Finally, the rate of reaction in the 60 seconds around the needle removal was 5·9/1000. The rate of VVS reaction remained generally higher in the incomplete group than the complete group for the first 4 minutes post needle withdrawal, varying from .06 to .28/1000 donations. With incomplete donations as with complete donations, Period 3 begins with a small peak in reactions (0·4/1000 donations). By the end of the 15 -minute period after phlebotomy, 98% of the VVS reactions had occurred. The last VVS associated with an incomplete donation occurred 106 minutes post needle withdrawal.

The female reaction rate was greater than the male reaction rate and women tended to have their reactions at a later time in relation to the donation than men (Fig. 3). The mean onset time for VVS in female donors (12·8 minutes, 95% CI 11·6–14·1) was significantly longer (P < 0·0001) than that in male donors (5·8 minutes, 95% CI, 4·7–6·9). During Period 2, 54% of the male VVS reactions but only 36% of the female VVS reactions associated with complete donations occurred. Six minutes after needle removal, 61% of all the male VVS reactions associated with complete donations had already occurred, but only 41% of all the female VVS reactions. Similarly, at the 6 -minute period, 100% of male VVS reactions had occurred, but only 94% of the female VVS reactions because of incomplete donations had occurred. At 15 minutes post needle withdrawal, 91% of the male VVS reactions because of complete donations had occurred, but only 75% of the female VVS reactions had occurred. Similarly at 15 minutes post needle withdrawal, 100% of the male VVS reactions to incomplete donations and 97% of the female VVS reactions to incomplete donations had occurred.


We chose to monitor VVS for this study because LOC is likely to be an objective outcome. It is problematic to compare reaction rates from one study to another because different reaction definitions are used and because definitions can be applied in different manners. While there still may be variability, the use of VVS should facilitate future comparison because there is less likelihood of variation in detecting loss-of-consciousness.

The duration of all donations is 8·23 minutes (95% CI 8·22–8·24); 7·66 (95% CI 7·65–7·68) and 8·62 (95% CI 8·61–8·63) for donations by men and women, respectively. Female donations lasted significantly longer than male donations. The same needle size was used for men and women. That this difference is clinically significant or significant in an analysis of VVS reactions should be the subject of future study.

There is a very low level of VVS in Period 1, the Period before venipunctures. These reactions cannot be associated with blood donation induced hypovolemia.

During Period 2, the venipuncture initiates a slow but steady increase in the reaction rate throughout phlebotomy. There is a sharp peak in reactions at the time the needle is removed. Two hundred and fifty-four (254) reactions occurred during the 60 seconds around the time the needle was removed, 104 reactions were associated with incomplete donations. For incomplete donations, this peak constitutes more than 60% of all reactions. For complete donations, this peak is only about 15% of all reactions. It is probable that for incomplete donations, the reaction “causes” the incomplete donation; i.e. the donation has to be stopped because the staff treats the reacting donor. For incomplete donations, the peak at needle removal is, at least partially, explainable. For complete donations, there is a peak of reactions at needle removal as well, though this peak is a much smaller percent of all VVS reactions. The explanation of this peak is not clear at present. Collection staff indicates some reactions begin while the staff is placing the blood unit in an area for further processing and these reactions are noted as “0-time reactions” when the staff member returns to the donor’s side. The reaction rate then decreases steadily during the remaining portion of Period 2.

Period 3 begins with the first of two peaks. The 5-minute and 9-minute peaks in Period 3 are smaller than the peak at needle removal, but the first peak is present for both male and female donors and for complete and incomplete reactions. It is likely that between 4 and 5 minutes post donation, donors begin to stand up. These peaks may represent the time when many donors stand up and some of these donors have reactions on assuming upright posture. After 9 minutes, there is a steady decrease in donor VVS reaction rates.

Forty per cent of the VVS reactions because of whole blood donation occur during Period 2. Period 2 was ∼14 minutes in length including the time from venipuncture until 4 minutes after needle withdrawal. Interventions aimed at this Period will be appropriate for ∼40% of all VVS reactions because of whole blood donation. Sixty per cent of the reactions occur in Period 3 which lasted 261 minutes in this study, from 4 minutes after the needle was removed until 265 minutes at which time the last VVS reaction was recorded. The duration of risk of VVS after phlebotomy was 265 minutes in our study, or 250 minutes from a time when most donors might leave the donation site. The rate was the highest during the time the donor is on-site with 80% of the VVS reactions occurring before the 15 minute time after needle withdrawal. The reactions occurring after the donor stands up are attributable at least in part to the donor’s inability to manage the relative hypovolemia that occurs after blood donation.

Diversion of the donor’s attention towards music, television or computer games, drinking water before donation and muscle tensing exercises have been tried to reduce the number VVS events. These techniques have been showed to be effective when applied to a correct segment of the donor population [3–9]. No firm data are currently available concerning the impact of these interventions on reactions which occur off-site and which are more likely to lead to donor injury.

In this study, we have presented the detailed time course of VVS reactions preceding and following whole blood donation from one year of collections by BSI. It is our hope that through further study it will be possible to associate the appropriate risk factors and levels of risk for each period associated with whole blood donation. We expect that future analyses and physiologic studies will refine our definitions of periods and assessment of duration of risk. It is also our intention that this set of observations will be useful to select interventions to reduce donor injury and to measure the impact of those interventions by comparing patterns of reactions post intervention.


While the data presented here were gathered at centres using the same standard operating procedure, there is known variation in the reaction rate among the centres [2]. The observations we have reported may not be the same at other blood centres. BSI did not note the time the donor stands up during this study, nor did BSI note the donor’s posture when the first symptoms of VVS were noted. Having these data would improve our analyses. There is no active surveillance after donors depart the donation site. We concluded that Period 3 was 261 minutes in duration, but the possibility of unreported reactions means that the duration of Period 3 could be longer. Our system of documentation and our SOP could produce some of the patterns we have observed. Broad conclusions from our data may not be appropriate until reproduced by other investigators.


Sixty per cent of the VVS reactions occurring in association with whole blood donation occur after the donor stands up (during Period 3). If the donor faints from an upright position, s/he is more likely to suffer injury than if the reaction occurs while the donor is recumbent. In addition, 8·7% of VVS reactions occur after the donor leaves the donation site and these reactions may be the most likely to lead to injury because the donor is no longer in the protective environment of the blood drive/clinic. Those donations occurring during Period 3 are associated with relative hypovolemia and interventions to prevent Period 3 VVS reactions should include effort to restore blood volume.