See all resources
Add to bookmarks
Article

SPI White Paper

Surgical Plethysmographic Index (SPI) in Anesthesia Practice

This paper explores answers to four questions about the Surgical Plethysmographic Index (SPI), based on data from clinical studies comprising of 35 sites and 1,496 patients.

Surgical Plethysmographic Index (SPI) in Anesthesia Practice

Matti Huiku, PhD., Lasse Kamppari, MSc. Hanna Viertio-Oja, PhD.

The purpose of this article is to provide answers to the following questions:

  1. How does SPI work?
  2. What is the information that SPI provides?
  3. Which physiological and external factors may affect the SPI values?
  4. What are the demonstrated clinical benefits of SPI monitoring?

Answers to these questions were collected from the clinical studies in Table 1 comprising of 35 clinical sites and 1496 patients.

The studies have been performed over the years 2005 - 2012.

The studies followed good clinical practice including ethical board approval and compliance with regulatory requirements for research in human subjects.

1. HOW DOES SPI WORK?

Painful stimulation causes sympathetic responses of the autonomic nervous system. During a surgical operation, such responses are normally suppressed by analgetic medication. If administration of analgesia is inadequate relative to the level of the stimulation, the patient may show responses such as increased heart rate and peripheral vasoconstriction. SPI derives both these components of information from the plethysmographic signal that is measured by a pulse oximeter. First both the heart rate and the plethysmographic amplitude are normalized in order to decrease inter-patient variability by applying a histogram transformation on the raw time serie data. Then a linear combination of the normalized values is computed as

SPI = 100-(0.7*PPGAnorm+0.3*HBInorm),

in which PPGAnorm is the normalized plethysmographic pulse wave amplitude and HBInorm the normalized heart beat interval. A detailed description of the method can be found in S1.

2. WHAT IS THE INFORMATION THAT SPI PROVIDES?

Study S1 was the first one to provide evidence of the two main properties of the SPI measurement:

1)     SPI values respond to surgical nociceptive stimuli, and
2)     SPI responses to stimulation are systematically modulated by analgesic drug concentration

Anesthesia in S1 was delivered using propofol and remifentanil. Since then the relationship of the SPI measurement to surgical stimulus and different types of analgesic medications has been addressed in eleven peer reviewed clinical study publications (S1, S2, S3, S4, S6, S8, S9, S11, S12, S13 and S20). The goal of these trials was to verify that the SPI reacts to the level of analgesic medication and surgical stimuli in a variety of analgesic, anesthetic and surgical regimens. The studies cover different clinical conditions, i.e. include various types of patients, anesthesia, analgesia, medications, and surgery. SPI shows statistically significant responses to surgical stimulation, which is evaluated by comparing the SPI values before and after incision and at known surgical noxious events. SPI also shows statistically significant responses to the changes of the level of analgesia (S1, S3, S4, S6, S11 and S13). In target controlled infusion the analgesic effect has been changed by altering the infusion rate and registering the change in SPI. In other types of analgesic regimen the SPI value has been registered before and after a bolus dose of analgesics.

Additionally, Struys et al (S3) has shown that while SPI is significantly lowered by analgesic medication (Remifentanil), it is not affected by non-analgesic medication (Propofol). This study result is supported by Ojala et al (A2). Ahonen et al (S2) has shown that SPI reflects the nociception-anti-nociception balance in patients receiving beta-blocking agent (Esmolol). Further, in this study the SPI reactivity seems not to be influenced by Esmolol.

The performance of the SPI measurement for assessing the level of nociception and analgesia has been compared to a measurement of skin conductivity by Ledowsky et al (S12). They concluded that Surgical Pleth Index, heart rate and blood pressure, but not the “number of fluctuations in skin conductance” changed in response to changes in level of analgesia by showing significant differences between before and after a bolus of Fentanyl.

It should be noted that SPI measures responses, and therefore it does not provide information of the effect of analgesic drug medication in the absence of stimulation. If there is no stimulation, there are no responses to measure, independently of the level of analgesia.

In summary, SPI provides information of the responses of the patient to ongoing surgical stimulation during general anesthesia and helps to optimize the level of analgesic medication according to the needs of an individual patient.

3. WHICH PHYSIOLOGICAL AND EXTERNAL FACTORS MAY AFFECT THE SPI VALUES?

Surgical Pleth Index reflects a change of the autonomic nervous system balance in body. The increase of the sympathetic activity increases SPI. Potentially, any medication or therapy that affects the sympathetic nervous system balance is reflected in the value of the SPI. The change in the autonomic balance, however, does not necessarily change the reactivity of SPI to analgesic medication or surgical stimulation. In fact, the reactivity is usually maintained, but the interpretation of the absolute SPI level is confounded.

Bonhomme and Hans et al (S15, S18) have studied SPI, HR and mean arterial pressure, MAP, during neurosurgery using standardized noxious stimulus and fluid challenge as interventions. They found that the interpretation of SPI, HR, and MAP is affected by the status of the patient’s intravascular blood volume and chronic history of hypertension. Based on the stimulus-response-probability model they suggested that the prediction for the adequate remifentanil level is the best when these factors were taken into account and the parameters were in concord in their prediction.

Mustola et al (S13) showed that Fentanyl boluses during operation decreased SPI and tracheal intubation, skin incision, and surgical stimuli during the procedure increased SPI as they should. SPI reacted consistently in both Sevoflurane and Isoflurane anesthesia. However, they noted that sometimes with poor plethysmographic signal in elderly patients at low blood pressure below MAP 60 mmHg and in the absence of stimulation SPI increased and NIBP decreased. This paradoxical response was reversed by Ethylphenylephrin 2 mg i.v.. Poor signal and weak plethysmographic pulse may thus confound the interpretation of the SPI values.

The study (S10) of Höcker et al investigated the effects of atropine and electric cardiac stimulation (pacemaker) on SPI in the absence of painful stimulation. The increase in HR induced by cardiac pacing or atropine resulted in a considerable increase in SPI without a change of the nociception- anti-nociception balance. They suggested that the absolute level of SPI should be interpreted cautiously in these situations.

Ilies et al (S16) investigated the effect of posture on SPI. The posture of a patient shifts the balance of the autonomic nervous system, and, therefore, has a marked effect on the absolute value, but does not suppress the reactivity of the Surgical Pleth Index.

The study of Ducrocq et al (A7) investigated the effect of ephedrine on SPI. Bolus doses of ephedrine were injected intravenously when systolic arterial pressure values were below 90 mmHg for more than 2 minutes. The study results showed that the ephedrine effect is transient and after 200 sec SPI is not any more influenced by ephedrine bolus. After the short transient the absolute SPI level can again be used as an estimator of the nociception - anti-nociception balance.

Aho et al (S4) investigated the effect of hypothermia on SPI. They found that the correlation to core temperature was only 0.022, indicating that mild hypothermia does not affect SPI. Further, the average SPI values were higher during than before surgery and higher with low than high remifentanil levels during surgery. The reactivity of SPI to surgical stimuli and opioid analgesia remained during moderate hypothermia.

Our supplementary studies, not in Table1, revealed that SPI is affected by cardiac arrhythmia, which especially when severe can cause unstable SPI values. In arrhythmia cases SPI should be interpreted cautiously.

SPI algorithm has been developed for finger measurement and cannot be used on other sites such as ear or toe. The change of the SpO₂ probe between fingers may also change SPI values, but will maintain the responsiveness of the measurement to stimulation and analgesic medications. As conclusion, after a change of the SpO₂ probe site or after removing and re-attaching the probe on the same site, the SPI value may shift to a slightly different level without a change in the nociception-anti-nociception balance.

In most cases the limitations of using SPI are obvious and relate to other than nociception and analgesic medication related changes in the patient’s pulse rate or plethysmographic amplitude (R1). Therefore, the interpretation of the SPI values, especially the absolute level of SPI, shall be done in conjunction to other physiological parameters, and therapy forms such as vasoactive and cardiac medications given to a patient.

4. WHAT ARE THE DEMONSTRATED CLINICAL BENEFITS OF SPI MONITORING?

Chen et al. (S14) compared SPI guided analgesia to standard clinical practice and concluded that SPI guided Remifentanil titration resulted in a significant reduction of Remifentanil consumption and less incidences of unwanted events such as hypertension, hypotension, tachycardia and movement during surgery (Figure 1.). Further, SPI showed the highest prediction probability for indicating maximum stimulation during surgery.

Number of unwanted events
Figure 1. Number of unwanted events during general anesthesia with Propofol and Remifentanil decreased by 85% in the SPI guided anesthesia group in comparison to the standard practice group (Chen et al, Comparison of Surgical Stress Index-guided Analgesia

Using the same patient population and study set-up Chen et al (S20) also studied the correlation of SPI with the level of stress hormones. They found that in intubation and during surgery the SPI could indicate the predefined specific level of ACTH, cortisol, epinephrine and norepinephrine with an area under ROC curve of 0.85, 0.62, 0.59 and 0.62, respectively. Sensitivity (81%) and specificity (73%) of the SPI to predict ACTH values were highest among the four stress hormones. They conclude that SPI was able to predict ACTH with high sensitivity and specificity.

The study by Bergmann (S17) compared the Propofol and Remifentanil consumption and operation process times between the SPI/Entropy guided anesthesia group and a control group with Entropy and standard clinical parameters in outpatient anesthesia. Both Remifentanil and Propofol consumption was significantly lower in the SPI/Entropy than in the Entropy group (p<0.05). The processing times for the opening of eyes and extubation was significantly lower in the SPI/Entropy than in the Entropy group (p<0.05). This study and the study by Chen et al (S14) show that the Propofol-Remifentanil anesthesia guided by SPI or SPI and Entropy may have an economic impact on patient care in terms of the consumption of anesthetics and the length of the operation.

In summary, the use of SPI during general anesthesia with Propofol and Remifentanil has been demonstrated to reduce the amount of adverse events, decrease the use of anesthetics, and shorten the time needed for the procedure.

APPENDIX: SUMMARY OF THE CLINICAL STUDIES ON SPI*

Clinical evidence for the Surgical Pleth Index, SPI, shown by the studies in Table 1 has been created in a broad range of surgical operations (Figure A1) and in all common types of general anesthesias (Figure A2 and A3). Most of the evaluation studies are for fully anesthetized adults (>=18 yrs) patients during general anesthesia (Figure A4), but SPI has also been evaluated for children and for spinal and local anesthesia with and without sedation, in awake patients in PACU and in chronic pain patients. Both known surgical stimulations and standardized electrical noxious stimulations are used to evaluate the performance of SPI (Figure A5). The usage of SPI for children and awake, sedated and lightly anesthetized patients are out of the intended use patient groups. The effect of certain, potentially confounding, common medications, and artifacts and surgical interventions has been evaluated, as well.

*For the Summary of clinical studies on SPI, please download the PDF attached

The number of patients in the SPI evaluation studies categorized based on the type of surgery. The total number of clinical cases with SPI is 1496.
Figure A1. The number of patients in the SPI evaluation studies categorized based on the type of surgery. The total number of clinical cases with SPI is 1496.
The number of patients in the SPI evaluations categorized by the different types of hypnotic agents.
Figure A2. The number of patients in the SPI evaluations categorized by the different types of hypnotic agents.
The number of patients in the SPI evaluations categorized by the different types of analgesia.
Figure A3. The number of patients in the SPI evaluations categorized by the different types of analgesia.
SPI evaluations per patient groups. SPI is intended for fully anesthetized adult patients (GA Adults).
Figure A4. SPI evaluations per patient groups. SPI is intended for fully anesthetized adult patients (GA Adults).
The performance of SPI has been evaluated at broad range of stimulation levels.
Figure A5. The performance of SPI has been evaluated at broad range of stimulation levels.

Propofol is both fast effecting and almost pure hypnotic agent and remifentanil has a very fast and the shortest time analgesic effect. Therefore, in a propofol-remifentanil anesthesia, the analgesic effect can be investigated separately from the hypnotic effect. Further, both the level of hypnosis and analgesia can be adjusted up and down in short time scales. Therefore, the responses of SPI can be demonstrated in a clear way and the interpretation of the expected effect of the medication is easy.

The inhalational anesthetic agents are typically used with opioids such as fentanyl or alfentanil that have a long lasting and slower effect time. Further, the effect of the inhalational anesthetic agent can be decreased only in relatively long time scales by increasing the ventilation and waiting until the end tidal concentration decreases. The gaseous anesthetic agents also have both hypnotic and analgesic effects in addition that the effects are highly synergistic. Correspondingly, the SPI changes are slower and less clear as they appear on top of the analgesic baseline provided by the inhalational agent alone. Consequently, the SPI responses can be studied between time points, i.e. intervals that are longer and may contain a change in the level of stimulation, which can be very dynamic in some surgical procedures. This smears out the SPI responses and make them more difficult to interpret in terms of the expected change in the balance of the noxious stimulation and analgesic medication.

The total number of patient cases in the studies is 1496. A broad range of surgical operations and all common types of general anesthesia are covered.

SPI has been evaluated for fully anesthetized adults (>=18 yrs) patients during general anesthesia. Both known surgical stimulations and standardized electrical noxious stimulations have been used to evaluate the performance of SPI. The SPI level and responses have been evaluated over wide range of analgesic regimens. SPI has also been evaluated for children during general anesthesia and awake, sedated and lightly anesthetized patients, which are out of the intended use patient groups. The effect of certain, potentially confounding, medications, and artifacts and surgical interventions has been evaluated.  Supplementary studies, which are unpublished, have evaluated SPI during cardiac surgeries and are summarized later in this section.

The results show that SPI reflects the nociception - anti-nociception balance during general anesthesia.

SPi is not available for sale in US and has not been 510(k) cleared. Not available in all countries. 

  • Neurology
  • SPI
  • Perioperative care
  • Clinical